Training in progress, step 8000, checkpoint
Browse files- checkpoint-8000/README.md +209 -0
- checkpoint-8000/adapter_config.json +48 -0
- checkpoint-8000/adapter_model.safetensors +3 -0
- checkpoint-8000/chat_template.jinja +47 -0
- checkpoint-8000/optimizer.pt +3 -0
- checkpoint-8000/processor_config.json +28 -0
- checkpoint-8000/rng_state.pth +3 -0
- checkpoint-8000/scheduler.pt +3 -0
- checkpoint-8000/tokenizer.json +3 -0
- checkpoint-8000/tokenizer_config.json +26 -0
- checkpoint-8000/trainer_state.json +2074 -0
- checkpoint-8000/training_args.bin +3 -0
checkpoint-8000/README.md
ADDED
|
@@ -0,0 +1,209 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
base_model: google/medgemma-1.5-4b-it
|
| 3 |
+
library_name: peft
|
| 4 |
+
pipeline_tag: text-generation
|
| 5 |
+
tags:
|
| 6 |
+
- base_model:adapter:google/medgemma-1.5-4b-it
|
| 7 |
+
- lora
|
| 8 |
+
- sft
|
| 9 |
+
- transformers
|
| 10 |
+
- trl
|
| 11 |
+
---
|
| 12 |
+
|
| 13 |
+
# Model Card for Model ID
|
| 14 |
+
|
| 15 |
+
<!-- Provide a quick summary of what the model is/does. -->
|
| 16 |
+
|
| 17 |
+
|
| 18 |
+
|
| 19 |
+
## Model Details
|
| 20 |
+
|
| 21 |
+
### Model Description
|
| 22 |
+
|
| 23 |
+
<!-- Provide a longer summary of what this model is. -->
|
| 24 |
+
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
- **Developed by:** [More Information Needed]
|
| 28 |
+
- **Funded by [optional]:** [More Information Needed]
|
| 29 |
+
- **Shared by [optional]:** [More Information Needed]
|
| 30 |
+
- **Model type:** [More Information Needed]
|
| 31 |
+
- **Language(s) (NLP):** [More Information Needed]
|
| 32 |
+
- **License:** [More Information Needed]
|
| 33 |
+
- **Finetuned from model [optional]:** [More Information Needed]
|
| 34 |
+
|
| 35 |
+
### Model Sources [optional]
|
| 36 |
+
|
| 37 |
+
<!-- Provide the basic links for the model. -->
|
| 38 |
+
|
| 39 |
+
- **Repository:** [More Information Needed]
|
| 40 |
+
- **Paper [optional]:** [More Information Needed]
|
| 41 |
+
- **Demo [optional]:** [More Information Needed]
|
| 42 |
+
|
| 43 |
+
## Uses
|
| 44 |
+
|
| 45 |
+
<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
|
| 46 |
+
|
| 47 |
+
### Direct Use
|
| 48 |
+
|
| 49 |
+
<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
|
| 50 |
+
|
| 51 |
+
[More Information Needed]
|
| 52 |
+
|
| 53 |
+
### Downstream Use [optional]
|
| 54 |
+
|
| 55 |
+
<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
|
| 56 |
+
|
| 57 |
+
[More Information Needed]
|
| 58 |
+
|
| 59 |
+
### Out-of-Scope Use
|
| 60 |
+
|
| 61 |
+
<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
|
| 62 |
+
|
| 63 |
+
[More Information Needed]
|
| 64 |
+
|
| 65 |
+
## Bias, Risks, and Limitations
|
| 66 |
+
|
| 67 |
+
<!-- This section is meant to convey both technical and sociotechnical limitations. -->
|
| 68 |
+
|
| 69 |
+
[More Information Needed]
|
| 70 |
+
|
| 71 |
+
### Recommendations
|
| 72 |
+
|
| 73 |
+
<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
|
| 74 |
+
|
| 75 |
+
Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
|
| 76 |
+
|
| 77 |
+
## How to Get Started with the Model
|
| 78 |
+
|
| 79 |
+
Use the code below to get started with the model.
|
| 80 |
+
|
| 81 |
+
[More Information Needed]
|
| 82 |
+
|
| 83 |
+
## Training Details
|
| 84 |
+
|
| 85 |
+
### Training Data
|
| 86 |
+
|
| 87 |
+
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
| 88 |
+
|
| 89 |
+
[More Information Needed]
|
| 90 |
+
|
| 91 |
+
### Training Procedure
|
| 92 |
+
|
| 93 |
+
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
| 94 |
+
|
| 95 |
+
#### Preprocessing [optional]
|
| 96 |
+
|
| 97 |
+
[More Information Needed]
|
| 98 |
+
|
| 99 |
+
|
| 100 |
+
#### Training Hyperparameters
|
| 101 |
+
|
| 102 |
+
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
| 103 |
+
|
| 104 |
+
#### Speeds, Sizes, Times [optional]
|
| 105 |
+
|
| 106 |
+
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
| 107 |
+
|
| 108 |
+
[More Information Needed]
|
| 109 |
+
|
| 110 |
+
## Evaluation
|
| 111 |
+
|
| 112 |
+
<!-- This section describes the evaluation protocols and provides the results. -->
|
| 113 |
+
|
| 114 |
+
### Testing Data, Factors & Metrics
|
| 115 |
+
|
| 116 |
+
#### Testing Data
|
| 117 |
+
|
| 118 |
+
<!-- This should link to a Dataset Card if possible. -->
|
| 119 |
+
|
| 120 |
+
[More Information Needed]
|
| 121 |
+
|
| 122 |
+
#### Factors
|
| 123 |
+
|
| 124 |
+
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 125 |
+
|
| 126 |
+
[More Information Needed]
|
| 127 |
+
|
| 128 |
+
#### Metrics
|
| 129 |
+
|
| 130 |
+
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 131 |
+
|
| 132 |
+
[More Information Needed]
|
| 133 |
+
|
| 134 |
+
### Results
|
| 135 |
+
|
| 136 |
+
[More Information Needed]
|
| 137 |
+
|
| 138 |
+
#### Summary
|
| 139 |
+
|
| 140 |
+
|
| 141 |
+
|
| 142 |
+
## Model Examination [optional]
|
| 143 |
+
|
| 144 |
+
<!-- Relevant interpretability work for the model goes here -->
|
| 145 |
+
|
| 146 |
+
[More Information Needed]
|
| 147 |
+
|
| 148 |
+
## Environmental Impact
|
| 149 |
+
|
| 150 |
+
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
| 151 |
+
|
| 152 |
+
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
| 153 |
+
|
| 154 |
+
- **Hardware Type:** [More Information Needed]
|
| 155 |
+
- **Hours used:** [More Information Needed]
|
| 156 |
+
- **Cloud Provider:** [More Information Needed]
|
| 157 |
+
- **Compute Region:** [More Information Needed]
|
| 158 |
+
- **Carbon Emitted:** [More Information Needed]
|
| 159 |
+
|
| 160 |
+
## Technical Specifications [optional]
|
| 161 |
+
|
| 162 |
+
### Model Architecture and Objective
|
| 163 |
+
|
| 164 |
+
[More Information Needed]
|
| 165 |
+
|
| 166 |
+
### Compute Infrastructure
|
| 167 |
+
|
| 168 |
+
[More Information Needed]
|
| 169 |
+
|
| 170 |
+
#### Hardware
|
| 171 |
+
|
| 172 |
+
[More Information Needed]
|
| 173 |
+
|
| 174 |
+
#### Software
|
| 175 |
+
|
| 176 |
+
[More Information Needed]
|
| 177 |
+
|
| 178 |
+
## Citation [optional]
|
| 179 |
+
|
| 180 |
+
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
| 181 |
+
|
| 182 |
+
**BibTeX:**
|
| 183 |
+
|
| 184 |
+
[More Information Needed]
|
| 185 |
+
|
| 186 |
+
**APA:**
|
| 187 |
+
|
| 188 |
+
[More Information Needed]
|
| 189 |
+
|
| 190 |
+
## Glossary [optional]
|
| 191 |
+
|
| 192 |
+
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
| 193 |
+
|
| 194 |
+
[More Information Needed]
|
| 195 |
+
|
| 196 |
+
## More Information [optional]
|
| 197 |
+
|
| 198 |
+
[More Information Needed]
|
| 199 |
+
|
| 200 |
+
## Model Card Authors [optional]
|
| 201 |
+
|
| 202 |
+
[More Information Needed]
|
| 203 |
+
|
| 204 |
+
## Model Card Contact
|
| 205 |
+
|
| 206 |
+
[More Information Needed]
|
| 207 |
+
### Framework versions
|
| 208 |
+
|
| 209 |
+
- PEFT 0.19.1
|
checkpoint-8000/adapter_config.json
ADDED
|
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"alora_invocation_tokens": null,
|
| 3 |
+
"alpha_pattern": {},
|
| 4 |
+
"arrow_config": null,
|
| 5 |
+
"auto_mapping": null,
|
| 6 |
+
"base_model_name_or_path": "google/medgemma-1.5-4b-it",
|
| 7 |
+
"bias": "none",
|
| 8 |
+
"corda_config": null,
|
| 9 |
+
"ensure_weight_tying": false,
|
| 10 |
+
"eva_config": null,
|
| 11 |
+
"exclude_modules": null,
|
| 12 |
+
"fan_in_fan_out": false,
|
| 13 |
+
"inference_mode": true,
|
| 14 |
+
"init_lora_weights": true,
|
| 15 |
+
"layer_replication": null,
|
| 16 |
+
"layers_pattern": null,
|
| 17 |
+
"layers_to_transform": null,
|
| 18 |
+
"loftq_config": {},
|
| 19 |
+
"lora_alpha": 16,
|
| 20 |
+
"lora_bias": false,
|
| 21 |
+
"lora_dropout": 0.05,
|
| 22 |
+
"lora_ga_config": null,
|
| 23 |
+
"megatron_config": null,
|
| 24 |
+
"megatron_core": "megatron.core",
|
| 25 |
+
"modules_to_save": null,
|
| 26 |
+
"peft_type": "LORA",
|
| 27 |
+
"peft_version": "0.19.1",
|
| 28 |
+
"qalora_group_size": 16,
|
| 29 |
+
"r": 16,
|
| 30 |
+
"rank_pattern": {},
|
| 31 |
+
"revision": null,
|
| 32 |
+
"target_modules": [
|
| 33 |
+
"down_proj",
|
| 34 |
+
"k_proj",
|
| 35 |
+
"o_proj",
|
| 36 |
+
"q_proj",
|
| 37 |
+
"v_proj",
|
| 38 |
+
"gate_proj",
|
| 39 |
+
"up_proj"
|
| 40 |
+
],
|
| 41 |
+
"target_parameters": null,
|
| 42 |
+
"task_type": "CAUSAL_LM",
|
| 43 |
+
"trainable_token_indices": null,
|
| 44 |
+
"use_bdlora": null,
|
| 45 |
+
"use_dora": false,
|
| 46 |
+
"use_qalora": false,
|
| 47 |
+
"use_rslora": false
|
| 48 |
+
}
|
checkpoint-8000/adapter_model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:74018f591fc3ef6a52dabf6a732d93017bf08b8c650639c4743979de3579bcd6
|
| 3 |
+
size 131250184
|
checkpoint-8000/chat_template.jinja
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{{ bos_token }}
|
| 2 |
+
{%- if messages[0]['role'] == 'system' -%}
|
| 3 |
+
{%- if messages[0]['content'] is string -%}
|
| 4 |
+
{%- set first_user_prefix = messages[0]['content'] + '
|
| 5 |
+
|
| 6 |
+
' -%}
|
| 7 |
+
{%- else -%}
|
| 8 |
+
{%- set first_user_prefix = messages[0]['content'][0]['text'] + '
|
| 9 |
+
|
| 10 |
+
' -%}
|
| 11 |
+
{%- endif -%}
|
| 12 |
+
{%- set loop_messages = messages[1:] -%}
|
| 13 |
+
{%- else -%}
|
| 14 |
+
{%- set first_user_prefix = "" -%}
|
| 15 |
+
{%- set loop_messages = messages -%}
|
| 16 |
+
{%- endif -%}
|
| 17 |
+
{%- for message in loop_messages -%}
|
| 18 |
+
{%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%}
|
| 19 |
+
{{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }}
|
| 20 |
+
{%- endif -%}
|
| 21 |
+
{%- if (message['role'] == 'assistant') -%}
|
| 22 |
+
{%- set role = "model" -%}
|
| 23 |
+
{%- else -%}
|
| 24 |
+
{%- set role = message['role'] -%}
|
| 25 |
+
{%- endif -%}
|
| 26 |
+
{{ '<start_of_turn>' + role + '
|
| 27 |
+
' + (first_user_prefix if loop.first else "") }}
|
| 28 |
+
{%- if message['content'] is string -%}
|
| 29 |
+
{{ message['content'] | trim }}
|
| 30 |
+
{%- elif message['content'] is iterable -%}
|
| 31 |
+
{%- for item in message['content'] -%}
|
| 32 |
+
{%- if item['type'] == 'image' -%}
|
| 33 |
+
{{ '<start_of_image>' }}
|
| 34 |
+
{%- elif item['type'] == 'text' -%}
|
| 35 |
+
{{ item['text'] | trim }}
|
| 36 |
+
{%- endif -%}
|
| 37 |
+
{%- endfor -%}
|
| 38 |
+
{%- else -%}
|
| 39 |
+
{{ raise_exception("Invalid content type") }}
|
| 40 |
+
{%- endif -%}
|
| 41 |
+
{{ '<end_of_turn>
|
| 42 |
+
' }}
|
| 43 |
+
{%- endfor -%}
|
| 44 |
+
{%- if add_generation_prompt -%}
|
| 45 |
+
{{'<start_of_turn>model
|
| 46 |
+
'}}
|
| 47 |
+
{%- endif -%}
|
checkpoint-8000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:202927e4b0dbb54c134f2e7cdeef4c31472445a4db8b392dde3122b2df0466c1
|
| 3 |
+
size 238828343
|
checkpoint-8000/processor_config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"image_processor": {
|
| 3 |
+
"do_convert_rgb": true,
|
| 4 |
+
"do_normalize": true,
|
| 5 |
+
"do_rescale": true,
|
| 6 |
+
"do_resize": true,
|
| 7 |
+
"image_mean": [
|
| 8 |
+
0.5,
|
| 9 |
+
0.5,
|
| 10 |
+
0.5
|
| 11 |
+
],
|
| 12 |
+
"image_processor_type": "Gemma3ImageProcessor",
|
| 13 |
+
"image_seq_length": 256,
|
| 14 |
+
"image_std": [
|
| 15 |
+
0.5,
|
| 16 |
+
0.5,
|
| 17 |
+
0.5
|
| 18 |
+
],
|
| 19 |
+
"resample": 2,
|
| 20 |
+
"rescale_factor": 0.00392156862745098,
|
| 21 |
+
"size": {
|
| 22 |
+
"height": 896,
|
| 23 |
+
"width": 896
|
| 24 |
+
}
|
| 25 |
+
},
|
| 26 |
+
"image_seq_length": 256,
|
| 27 |
+
"processor_class": "Gemma3Processor"
|
| 28 |
+
}
|
checkpoint-8000/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7b4abe3f54819bae8961acad0d167300be9bdfc72ea26403ade54439414d7185
|
| 3 |
+
size 14645
|
checkpoint-8000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fee5cf1570360d6f68ad929f99b86c0a66c93513fe5867e46ddf496b40a4cc02
|
| 3 |
+
size 1465
|
checkpoint-8000/tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a6c776bcef3fcea90b4244ef4dafa43ec828ea9f23a286a5be6bd3c209efe810
|
| 3 |
+
size 33384721
|
checkpoint-8000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"backend": "pil",
|
| 3 |
+
"boi_token": "<start_of_image>",
|
| 4 |
+
"bos_token": "<bos>",
|
| 5 |
+
"clean_up_tokenization_spaces": false,
|
| 6 |
+
"eoi_token": "<end_of_image>",
|
| 7 |
+
"eos_token": "<eos>",
|
| 8 |
+
"image_token": "<image_soft_token>",
|
| 9 |
+
"is_local": false,
|
| 10 |
+
"local_files_only": false,
|
| 11 |
+
"mask_token": "<mask>",
|
| 12 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 13 |
+
"model_specific_special_tokens": {
|
| 14 |
+
"boi_token": "<start_of_image>",
|
| 15 |
+
"eoi_token": "<end_of_image>",
|
| 16 |
+
"image_token": "<image_soft_token>"
|
| 17 |
+
},
|
| 18 |
+
"pad_token": "<pad>",
|
| 19 |
+
"padding_side": "left",
|
| 20 |
+
"processor_class": "Gemma3Processor",
|
| 21 |
+
"sp_model_kwargs": null,
|
| 22 |
+
"spaces_between_special_tokens": false,
|
| 23 |
+
"tokenizer_class": "GemmaTokenizer",
|
| 24 |
+
"unk_token": "<unk>",
|
| 25 |
+
"use_default_system_prompt": false
|
| 26 |
+
}
|
checkpoint-8000/trainer_state.json
ADDED
|
@@ -0,0 +1,2074 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 6800,
|
| 3 |
+
"best_metric": 0.10720802843570709,
|
| 4 |
+
"best_model_checkpoint": "data/full_sft_run/sft_out/checkpoint-6800",
|
| 5 |
+
"epoch": 2.337888198757764,
|
| 6 |
+
"eval_steps": 200,
|
| 7 |
+
"global_step": 8000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"entropy": 0.20752533283084632,
|
| 14 |
+
"epoch": 0.014614541468761417,
|
| 15 |
+
"grad_norm": 0.7449454069137573,
|
| 16 |
+
"learning_rate": 3.181818181818182e-05,
|
| 17 |
+
"loss": 1.3096963500976562,
|
| 18 |
+
"mean_token_accuracy": 0.8356087231636047,
|
| 19 |
+
"num_tokens": 218234.0,
|
| 20 |
+
"step": 50
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"entropy": 0.2586296512931585,
|
| 24 |
+
"epoch": 0.029229082937522834,
|
| 25 |
+
"grad_norm": 0.6199880838394165,
|
| 26 |
+
"learning_rate": 6.428571428571429e-05,
|
| 27 |
+
"loss": 0.26672126770019533,
|
| 28 |
+
"mean_token_accuracy": 0.9052031901478768,
|
| 29 |
+
"num_tokens": 434925.0,
|
| 30 |
+
"step": 100
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"entropy": 0.2025177276879549,
|
| 34 |
+
"epoch": 0.043843624406284254,
|
| 35 |
+
"grad_norm": 0.5352309346199036,
|
| 36 |
+
"learning_rate": 9.675324675324677e-05,
|
| 37 |
+
"loss": 0.20794065475463866,
|
| 38 |
+
"mean_token_accuracy": 0.9208106356859207,
|
| 39 |
+
"num_tokens": 652823.0,
|
| 40 |
+
"step": 150
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"entropy": 0.18583494622260333,
|
| 44 |
+
"epoch": 0.05845816587504567,
|
| 45 |
+
"grad_norm": 0.4790455996990204,
|
| 46 |
+
"learning_rate": 0.00012922077922077922,
|
| 47 |
+
"loss": 0.19879446029663086,
|
| 48 |
+
"mean_token_accuracy": 0.9228935891389847,
|
| 49 |
+
"num_tokens": 870027.0,
|
| 50 |
+
"step": 200
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.05845816587504567,
|
| 54 |
+
"eval_entropy": 0.20352977880481946,
|
| 55 |
+
"eval_loss": 0.1885823905467987,
|
| 56 |
+
"eval_mean_token_accuracy": 0.9265667232416444,
|
| 57 |
+
"eval_num_tokens": 870027.0,
|
| 58 |
+
"eval_runtime": 51.9934,
|
| 59 |
+
"eval_samples_per_second": 22.657,
|
| 60 |
+
"eval_steps_per_second": 5.674,
|
| 61 |
+
"step": 200
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"entropy": 0.18395513169467448,
|
| 65 |
+
"epoch": 0.0730727073438071,
|
| 66 |
+
"grad_norm": 0.5151433348655701,
|
| 67 |
+
"learning_rate": 0.0001616883116883117,
|
| 68 |
+
"loss": 0.1874466896057129,
|
| 69 |
+
"mean_token_accuracy": 0.9266703170537949,
|
| 70 |
+
"num_tokens": 1086768.0,
|
| 71 |
+
"step": 250
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"entropy": 0.18530432548373937,
|
| 75 |
+
"epoch": 0.08768724881256851,
|
| 76 |
+
"grad_norm": 0.360085666179657,
|
| 77 |
+
"learning_rate": 0.00019415584415584416,
|
| 78 |
+
"loss": 0.18927602767944335,
|
| 79 |
+
"mean_token_accuracy": 0.9263237661123276,
|
| 80 |
+
"num_tokens": 1306053.0,
|
| 81 |
+
"step": 300
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"entropy": 0.18520764522254468,
|
| 85 |
+
"epoch": 0.10230179028132992,
|
| 86 |
+
"grad_norm": 0.32682710886001587,
|
| 87 |
+
"learning_rate": 0.00019917654147419162,
|
| 88 |
+
"loss": 0.19010570526123047,
|
| 89 |
+
"mean_token_accuracy": 0.9269409370422363,
|
| 90 |
+
"num_tokens": 1524650.0,
|
| 91 |
+
"step": 350
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"entropy": 0.17121127102524042,
|
| 95 |
+
"epoch": 0.11691633175009133,
|
| 96 |
+
"grad_norm": 0.33935022354125977,
|
| 97 |
+
"learning_rate": 0.00019817232375979112,
|
| 98 |
+
"loss": 0.17950326919555665,
|
| 99 |
+
"mean_token_accuracy": 0.928639045059681,
|
| 100 |
+
"num_tokens": 1740391.0,
|
| 101 |
+
"step": 400
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.11691633175009133,
|
| 105 |
+
"eval_entropy": 0.15643637586953277,
|
| 106 |
+
"eval_loss": 0.17647278308868408,
|
| 107 |
+
"eval_mean_token_accuracy": 0.9313437746743024,
|
| 108 |
+
"eval_num_tokens": 1740391.0,
|
| 109 |
+
"eval_runtime": 51.4028,
|
| 110 |
+
"eval_samples_per_second": 22.917,
|
| 111 |
+
"eval_steps_per_second": 5.739,
|
| 112 |
+
"step": 400
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"entropy": 0.17545538306236266,
|
| 116 |
+
"epoch": 0.13153087321885276,
|
| 117 |
+
"grad_norm": 0.35228046774864197,
|
| 118 |
+
"learning_rate": 0.00019716810604539066,
|
| 119 |
+
"loss": 0.18172544479370117,
|
| 120 |
+
"mean_token_accuracy": 0.9291265362501144,
|
| 121 |
+
"num_tokens": 1957727.0,
|
| 122 |
+
"step": 450
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"entropy": 0.1679905788227916,
|
| 126 |
+
"epoch": 0.1461454146876142,
|
| 127 |
+
"grad_norm": 0.36982566118240356,
|
| 128 |
+
"learning_rate": 0.0001961638883309902,
|
| 129 |
+
"loss": 0.17352703094482422,
|
| 130 |
+
"mean_token_accuracy": 0.9319902381300926,
|
| 131 |
+
"num_tokens": 2174741.0,
|
| 132 |
+
"step": 500
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"entropy": 0.16595628045499325,
|
| 136 |
+
"epoch": 0.1607599561563756,
|
| 137 |
+
"grad_norm": 0.2658090889453888,
|
| 138 |
+
"learning_rate": 0.0001951596706165897,
|
| 139 |
+
"loss": 0.1694266128540039,
|
| 140 |
+
"mean_token_accuracy": 0.9349100396037102,
|
| 141 |
+
"num_tokens": 2392020.0,
|
| 142 |
+
"step": 550
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"entropy": 0.16491649385541676,
|
| 146 |
+
"epoch": 0.17537449762513702,
|
| 147 |
+
"grad_norm": 0.28709471225738525,
|
| 148 |
+
"learning_rate": 0.0001941554529021892,
|
| 149 |
+
"loss": 0.16749130249023436,
|
| 150 |
+
"mean_token_accuracy": 0.9363178312778473,
|
| 151 |
+
"num_tokens": 2609502.0,
|
| 152 |
+
"step": 600
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.17537449762513702,
|
| 156 |
+
"eval_entropy": 0.15555994159589379,
|
| 157 |
+
"eval_loss": 0.16351227462291718,
|
| 158 |
+
"eval_mean_token_accuracy": 0.935047954623982,
|
| 159 |
+
"eval_num_tokens": 2609502.0,
|
| 160 |
+
"eval_runtime": 51.7247,
|
| 161 |
+
"eval_samples_per_second": 22.774,
|
| 162 |
+
"eval_steps_per_second": 5.703,
|
| 163 |
+
"step": 600
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"entropy": 0.16599890314042567,
|
| 167 |
+
"epoch": 0.18998903909389842,
|
| 168 |
+
"grad_norm": 0.3486684560775757,
|
| 169 |
+
"learning_rate": 0.00019315123518778873,
|
| 170 |
+
"loss": 0.16908374786376953,
|
| 171 |
+
"mean_token_accuracy": 0.9338638958334923,
|
| 172 |
+
"num_tokens": 2826468.0,
|
| 173 |
+
"step": 650
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"entropy": 0.15440633669495582,
|
| 177 |
+
"epoch": 0.20460358056265984,
|
| 178 |
+
"grad_norm": 0.3269498944282532,
|
| 179 |
+
"learning_rate": 0.00019214701747338823,
|
| 180 |
+
"loss": 0.1564903450012207,
|
| 181 |
+
"mean_token_accuracy": 0.9378401437401771,
|
| 182 |
+
"num_tokens": 3044037.0,
|
| 183 |
+
"step": 700
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"entropy": 0.15702761143445967,
|
| 187 |
+
"epoch": 0.21921812203142127,
|
| 188 |
+
"grad_norm": 0.2996591329574585,
|
| 189 |
+
"learning_rate": 0.00019114279975898776,
|
| 190 |
+
"loss": 0.16098966598510742,
|
| 191 |
+
"mean_token_accuracy": 0.937047846019268,
|
| 192 |
+
"num_tokens": 3260671.0,
|
| 193 |
+
"step": 750
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"entropy": 0.16964259944856167,
|
| 197 |
+
"epoch": 0.23383266350018267,
|
| 198 |
+
"grad_norm": 0.27347952127456665,
|
| 199 |
+
"learning_rate": 0.00019013858204458727,
|
| 200 |
+
"loss": 0.17197874069213867,
|
| 201 |
+
"mean_token_accuracy": 0.9319348493218422,
|
| 202 |
+
"num_tokens": 3479952.0,
|
| 203 |
+
"step": 800
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 0.23383266350018267,
|
| 207 |
+
"eval_entropy": 0.15915015558317555,
|
| 208 |
+
"eval_loss": 0.15902064740657806,
|
| 209 |
+
"eval_mean_token_accuracy": 0.9373329009039927,
|
| 210 |
+
"eval_num_tokens": 3479952.0,
|
| 211 |
+
"eval_runtime": 52.0586,
|
| 212 |
+
"eval_samples_per_second": 22.628,
|
| 213 |
+
"eval_steps_per_second": 5.667,
|
| 214 |
+
"step": 800
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"entropy": 0.16319214086979628,
|
| 218 |
+
"epoch": 0.2484472049689441,
|
| 219 |
+
"grad_norm": 0.2520512342453003,
|
| 220 |
+
"learning_rate": 0.0001891343643301868,
|
| 221 |
+
"loss": 0.16472122192382813,
|
| 222 |
+
"mean_token_accuracy": 0.9351362636685372,
|
| 223 |
+
"num_tokens": 3695714.0,
|
| 224 |
+
"step": 850
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"entropy": 0.157652658931911,
|
| 228 |
+
"epoch": 0.2630617464377055,
|
| 229 |
+
"grad_norm": 0.2863970398902893,
|
| 230 |
+
"learning_rate": 0.00018813014661578633,
|
| 231 |
+
"loss": 0.15824624061584472,
|
| 232 |
+
"mean_token_accuracy": 0.9369783601164818,
|
| 233 |
+
"num_tokens": 3913429.0,
|
| 234 |
+
"step": 900
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"entropy": 0.15214558444917203,
|
| 238 |
+
"epoch": 0.2776762879064669,
|
| 239 |
+
"grad_norm": 0.290787935256958,
|
| 240 |
+
"learning_rate": 0.00018712592890138583,
|
| 241 |
+
"loss": 0.15539746284484862,
|
| 242 |
+
"mean_token_accuracy": 0.9397245013713836,
|
| 243 |
+
"num_tokens": 4131640.0,
|
| 244 |
+
"step": 950
|
| 245 |
+
},
|
| 246 |
+
{
|
| 247 |
+
"entropy": 0.15566726494580507,
|
| 248 |
+
"epoch": 0.2922908293752284,
|
| 249 |
+
"grad_norm": 0.3074010908603668,
|
| 250 |
+
"learning_rate": 0.00018612171118698534,
|
| 251 |
+
"loss": 0.15537342071533203,
|
| 252 |
+
"mean_token_accuracy": 0.9387755367159843,
|
| 253 |
+
"num_tokens": 4348832.0,
|
| 254 |
+
"step": 1000
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.2922908293752284,
|
| 258 |
+
"eval_entropy": 0.1574143134436365,
|
| 259 |
+
"eval_loss": 0.15738964080810547,
|
| 260 |
+
"eval_mean_token_accuracy": 0.9374752864999286,
|
| 261 |
+
"eval_num_tokens": 4348832.0,
|
| 262 |
+
"eval_runtime": 51.5665,
|
| 263 |
+
"eval_samples_per_second": 22.844,
|
| 264 |
+
"eval_steps_per_second": 5.721,
|
| 265 |
+
"step": 1000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"entropy": 0.1583068972080946,
|
| 269 |
+
"epoch": 0.3069053708439898,
|
| 270 |
+
"grad_norm": 0.3427266776561737,
|
| 271 |
+
"learning_rate": 0.00018511749347258487,
|
| 272 |
+
"loss": 0.15989484786987304,
|
| 273 |
+
"mean_token_accuracy": 0.9373269161581993,
|
| 274 |
+
"num_tokens": 4565702.0,
|
| 275 |
+
"step": 1050
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"entropy": 0.1511041523516178,
|
| 279 |
+
"epoch": 0.3215199123127512,
|
| 280 |
+
"grad_norm": 0.32759514451026917,
|
| 281 |
+
"learning_rate": 0.00018411327575818437,
|
| 282 |
+
"loss": 0.15331138610839845,
|
| 283 |
+
"mean_token_accuracy": 0.9397411552071572,
|
| 284 |
+
"num_tokens": 4784167.0,
|
| 285 |
+
"step": 1100
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"entropy": 0.14996831517666578,
|
| 289 |
+
"epoch": 0.33613445378151263,
|
| 290 |
+
"grad_norm": 0.30234530568122864,
|
| 291 |
+
"learning_rate": 0.0001831090580437839,
|
| 292 |
+
"loss": 0.1544932556152344,
|
| 293 |
+
"mean_token_accuracy": 0.9399288591742515,
|
| 294 |
+
"num_tokens": 4999793.0,
|
| 295 |
+
"step": 1150
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"entropy": 0.15582640528678893,
|
| 299 |
+
"epoch": 0.35074899525027403,
|
| 300 |
+
"grad_norm": 0.24690158665180206,
|
| 301 |
+
"learning_rate": 0.0001821048403293834,
|
| 302 |
+
"loss": 0.15442598342895508,
|
| 303 |
+
"mean_token_accuracy": 0.9398289120197296,
|
| 304 |
+
"num_tokens": 5217954.0,
|
| 305 |
+
"step": 1200
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.35074899525027403,
|
| 309 |
+
"eval_entropy": 0.1420629285275936,
|
| 310 |
+
"eval_loss": 0.14971356093883514,
|
| 311 |
+
"eval_mean_token_accuracy": 0.9414400791717787,
|
| 312 |
+
"eval_num_tokens": 5217954.0,
|
| 313 |
+
"eval_runtime": 51.4751,
|
| 314 |
+
"eval_samples_per_second": 22.885,
|
| 315 |
+
"eval_steps_per_second": 5.731,
|
| 316 |
+
"step": 1200
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"entropy": 0.1543054273724556,
|
| 320 |
+
"epoch": 0.36536353671903543,
|
| 321 |
+
"grad_norm": 0.29708531498908997,
|
| 322 |
+
"learning_rate": 0.00018110062261498294,
|
| 323 |
+
"loss": 0.1537049388885498,
|
| 324 |
+
"mean_token_accuracy": 0.9416613772511482,
|
| 325 |
+
"num_tokens": 5434380.0,
|
| 326 |
+
"step": 1250
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"entropy": 0.14713672876358033,
|
| 330 |
+
"epoch": 0.37997807818779683,
|
| 331 |
+
"grad_norm": 0.2888696789741516,
|
| 332 |
+
"learning_rate": 0.00018009640490058247,
|
| 333 |
+
"loss": 0.15287999153137208,
|
| 334 |
+
"mean_token_accuracy": 0.9416247588396073,
|
| 335 |
+
"num_tokens": 5651784.0,
|
| 336 |
+
"step": 1300
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"entropy": 0.14809815444052218,
|
| 340 |
+
"epoch": 0.3945926196565583,
|
| 341 |
+
"grad_norm": 0.25226420164108276,
|
| 342 |
+
"learning_rate": 0.00017909218718618197,
|
| 343 |
+
"loss": 0.15003498077392577,
|
| 344 |
+
"mean_token_accuracy": 0.9424466270208359,
|
| 345 |
+
"num_tokens": 5870443.0,
|
| 346 |
+
"step": 1350
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"entropy": 0.15500445548444985,
|
| 350 |
+
"epoch": 0.4092071611253197,
|
| 351 |
+
"grad_norm": 0.2410505712032318,
|
| 352 |
+
"learning_rate": 0.0001780879694717815,
|
| 353 |
+
"loss": 0.15515862464904784,
|
| 354 |
+
"mean_token_accuracy": 0.9403725653886795,
|
| 355 |
+
"num_tokens": 6086417.0,
|
| 356 |
+
"step": 1400
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.4092071611253197,
|
| 360 |
+
"eval_entropy": 0.13966038472066492,
|
| 361 |
+
"eval_loss": 0.1482466757297516,
|
| 362 |
+
"eval_mean_token_accuracy": 0.9430661270173929,
|
| 363 |
+
"eval_num_tokens": 6086417.0,
|
| 364 |
+
"eval_runtime": 51.3223,
|
| 365 |
+
"eval_samples_per_second": 22.953,
|
| 366 |
+
"eval_steps_per_second": 5.748,
|
| 367 |
+
"step": 1400
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"entropy": 0.14597053304314614,
|
| 371 |
+
"epoch": 0.4238217025940811,
|
| 372 |
+
"grad_norm": 0.2526480257511139,
|
| 373 |
+
"learning_rate": 0.000177083751757381,
|
| 374 |
+
"loss": 0.1495055389404297,
|
| 375 |
+
"mean_token_accuracy": 0.9429304957389831,
|
| 376 |
+
"num_tokens": 6304323.0,
|
| 377 |
+
"step": 1450
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"entropy": 0.1473805220425129,
|
| 381 |
+
"epoch": 0.43843624406284254,
|
| 382 |
+
"grad_norm": 0.27721503376960754,
|
| 383 |
+
"learning_rate": 0.0001760795340429805,
|
| 384 |
+
"loss": 0.14802109718322753,
|
| 385 |
+
"mean_token_accuracy": 0.9423278120160102,
|
| 386 |
+
"num_tokens": 6521073.0,
|
| 387 |
+
"step": 1500
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"entropy": 0.15090686846524476,
|
| 391 |
+
"epoch": 0.45305078553160394,
|
| 392 |
+
"grad_norm": 0.3168330192565918,
|
| 393 |
+
"learning_rate": 0.00017507531632858004,
|
| 394 |
+
"loss": 0.1520420265197754,
|
| 395 |
+
"mean_token_accuracy": 0.9409908491373062,
|
| 396 |
+
"num_tokens": 6738178.0,
|
| 397 |
+
"step": 1550
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"entropy": 0.14193191647529602,
|
| 401 |
+
"epoch": 0.46766532700036534,
|
| 402 |
+
"grad_norm": 0.28218454122543335,
|
| 403 |
+
"learning_rate": 0.00017407109861417957,
|
| 404 |
+
"loss": 0.1442671012878418,
|
| 405 |
+
"mean_token_accuracy": 0.944555883705616,
|
| 406 |
+
"num_tokens": 6955413.0,
|
| 407 |
+
"step": 1600
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 0.46766532700036534,
|
| 411 |
+
"eval_entropy": 0.14087487072762797,
|
| 412 |
+
"eval_loss": 0.14355963468551636,
|
| 413 |
+
"eval_mean_token_accuracy": 0.943681773694895,
|
| 414 |
+
"eval_num_tokens": 6955413.0,
|
| 415 |
+
"eval_runtime": 52.1208,
|
| 416 |
+
"eval_samples_per_second": 22.601,
|
| 417 |
+
"eval_steps_per_second": 5.66,
|
| 418 |
+
"step": 1600
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"entropy": 0.1416533422470093,
|
| 422 |
+
"epoch": 0.4822798684691268,
|
| 423 |
+
"grad_norm": 0.28569167852401733,
|
| 424 |
+
"learning_rate": 0.00017306688089977908,
|
| 425 |
+
"loss": 0.146381254196167,
|
| 426 |
+
"mean_token_accuracy": 0.9433900326490402,
|
| 427 |
+
"num_tokens": 7171960.0,
|
| 428 |
+
"step": 1650
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"entropy": 0.14460256807506083,
|
| 432 |
+
"epoch": 0.4968944099378882,
|
| 433 |
+
"grad_norm": 0.298966646194458,
|
| 434 |
+
"learning_rate": 0.0001720626631853786,
|
| 435 |
+
"loss": 0.14527856826782226,
|
| 436 |
+
"mean_token_accuracy": 0.9446076577901841,
|
| 437 |
+
"num_tokens": 7387948.0,
|
| 438 |
+
"step": 1700
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"entropy": 0.14507481291890145,
|
| 442 |
+
"epoch": 0.5115089514066496,
|
| 443 |
+
"grad_norm": 0.32264840602874756,
|
| 444 |
+
"learning_rate": 0.0001710584454709781,
|
| 445 |
+
"loss": 0.14816983222961425,
|
| 446 |
+
"mean_token_accuracy": 0.9431060332059861,
|
| 447 |
+
"num_tokens": 7605190.0,
|
| 448 |
+
"step": 1750
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"entropy": 0.1457299491018057,
|
| 452 |
+
"epoch": 0.526123492875411,
|
| 453 |
+
"grad_norm": 0.23384633660316467,
|
| 454 |
+
"learning_rate": 0.00017005422775657764,
|
| 455 |
+
"loss": 0.14449834823608398,
|
| 456 |
+
"mean_token_accuracy": 0.9444263905286789,
|
| 457 |
+
"num_tokens": 7822465.0,
|
| 458 |
+
"step": 1800
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.526123492875411,
|
| 462 |
+
"eval_entropy": 0.13727282429650678,
|
| 463 |
+
"eval_loss": 0.14151474833488464,
|
| 464 |
+
"eval_mean_token_accuracy": 0.9445956945419312,
|
| 465 |
+
"eval_num_tokens": 7822465.0,
|
| 466 |
+
"eval_runtime": 51.7253,
|
| 467 |
+
"eval_samples_per_second": 22.774,
|
| 468 |
+
"eval_steps_per_second": 5.703,
|
| 469 |
+
"step": 1800
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"entropy": 0.13303045216947795,
|
| 473 |
+
"epoch": 0.5407380343441724,
|
| 474 |
+
"grad_norm": 0.24926069378852844,
|
| 475 |
+
"learning_rate": 0.00016905001004217715,
|
| 476 |
+
"loss": 0.1377289867401123,
|
| 477 |
+
"mean_token_accuracy": 0.9460858461260796,
|
| 478 |
+
"num_tokens": 8039013.0,
|
| 479 |
+
"step": 1850
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"entropy": 0.14333279909566043,
|
| 483 |
+
"epoch": 0.5553525758129338,
|
| 484 |
+
"grad_norm": 0.23103737831115723,
|
| 485 |
+
"learning_rate": 0.00016804579232777665,
|
| 486 |
+
"loss": 0.14504934310913087,
|
| 487 |
+
"mean_token_accuracy": 0.9442342269420624,
|
| 488 |
+
"num_tokens": 8255914.0,
|
| 489 |
+
"step": 1900
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"entropy": 0.14443909596651794,
|
| 493 |
+
"epoch": 0.5699671172816952,
|
| 494 |
+
"grad_norm": 0.3567982017993927,
|
| 495 |
+
"learning_rate": 0.00016704157461337618,
|
| 496 |
+
"loss": 0.14560423851013182,
|
| 497 |
+
"mean_token_accuracy": 0.9445012074708938,
|
| 498 |
+
"num_tokens": 8473895.0,
|
| 499 |
+
"step": 1950
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"entropy": 0.1381967855617404,
|
| 503 |
+
"epoch": 0.5845816587504568,
|
| 504 |
+
"grad_norm": 0.32823431491851807,
|
| 505 |
+
"learning_rate": 0.00016603735689897571,
|
| 506 |
+
"loss": 0.1393825626373291,
|
| 507 |
+
"mean_token_accuracy": 0.9458042901754379,
|
| 508 |
+
"num_tokens": 8690259.0,
|
| 509 |
+
"step": 2000
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.5845816587504568,
|
| 513 |
+
"eval_entropy": 0.12425902821755005,
|
| 514 |
+
"eval_loss": 0.14112941920757294,
|
| 515 |
+
"eval_mean_token_accuracy": 0.9466315289675179,
|
| 516 |
+
"eval_num_tokens": 8690259.0,
|
| 517 |
+
"eval_runtime": 51.7824,
|
| 518 |
+
"eval_samples_per_second": 22.749,
|
| 519 |
+
"eval_steps_per_second": 5.697,
|
| 520 |
+
"step": 2000
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"entropy": 0.14191570974886417,
|
| 524 |
+
"epoch": 0.5991962002192182,
|
| 525 |
+
"grad_norm": 0.27820655703544617,
|
| 526 |
+
"learning_rate": 0.00016503313918457522,
|
| 527 |
+
"loss": 0.14510326385498046,
|
| 528 |
+
"mean_token_accuracy": 0.9463699841499329,
|
| 529 |
+
"num_tokens": 8908010.0,
|
| 530 |
+
"step": 2050
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"entropy": 0.14415166847407818,
|
| 534 |
+
"epoch": 0.6138107416879796,
|
| 535 |
+
"grad_norm": 0.2561301290988922,
|
| 536 |
+
"learning_rate": 0.00016402892147017475,
|
| 537 |
+
"loss": 0.144957275390625,
|
| 538 |
+
"mean_token_accuracy": 0.945547785460949,
|
| 539 |
+
"num_tokens": 9124790.0,
|
| 540 |
+
"step": 2100
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"entropy": 0.13635289475321768,
|
| 544 |
+
"epoch": 0.628425283156741,
|
| 545 |
+
"grad_norm": 0.2438308447599411,
|
| 546 |
+
"learning_rate": 0.00016302470375577425,
|
| 547 |
+
"loss": 0.13744013786315917,
|
| 548 |
+
"mean_token_accuracy": 0.9481257164478302,
|
| 549 |
+
"num_tokens": 9343016.0,
|
| 550 |
+
"step": 2150
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"entropy": 0.12979858005419373,
|
| 554 |
+
"epoch": 0.6430398246255024,
|
| 555 |
+
"grad_norm": 0.32355841994285583,
|
| 556 |
+
"learning_rate": 0.00016202048604137379,
|
| 557 |
+
"loss": 0.13397377014160156,
|
| 558 |
+
"mean_token_accuracy": 0.9486737114191055,
|
| 559 |
+
"num_tokens": 9559022.0,
|
| 560 |
+
"step": 2200
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.6430398246255024,
|
| 564 |
+
"eval_entropy": 0.12423726801397436,
|
| 565 |
+
"eval_loss": 0.1396605670452118,
|
| 566 |
+
"eval_mean_token_accuracy": 0.947921950938338,
|
| 567 |
+
"eval_num_tokens": 9559022.0,
|
| 568 |
+
"eval_runtime": 51.4931,
|
| 569 |
+
"eval_samples_per_second": 22.877,
|
| 570 |
+
"eval_steps_per_second": 5.729,
|
| 571 |
+
"step": 2200
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"entropy": 0.1332092508673668,
|
| 575 |
+
"epoch": 0.6576543660942638,
|
| 576 |
+
"grad_norm": 0.2890651524066925,
|
| 577 |
+
"learning_rate": 0.0001610162683269733,
|
| 578 |
+
"loss": 0.13556601524353026,
|
| 579 |
+
"mean_token_accuracy": 0.9473348143696785,
|
| 580 |
+
"num_tokens": 9774504.0,
|
| 581 |
+
"step": 2250
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"entropy": 0.13480757348239422,
|
| 585 |
+
"epoch": 0.6722689075630253,
|
| 586 |
+
"grad_norm": 0.23138317465782166,
|
| 587 |
+
"learning_rate": 0.0001600120506125728,
|
| 588 |
+
"loss": 0.13720419883728027,
|
| 589 |
+
"mean_token_accuracy": 0.9482911565899849,
|
| 590 |
+
"num_tokens": 9993556.0,
|
| 591 |
+
"step": 2300
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"entropy": 0.13804105442017317,
|
| 595 |
+
"epoch": 0.6868834490317867,
|
| 596 |
+
"grad_norm": 0.2838476896286011,
|
| 597 |
+
"learning_rate": 0.00015900783289817232,
|
| 598 |
+
"loss": 0.13747754096984863,
|
| 599 |
+
"mean_token_accuracy": 0.9490453228354454,
|
| 600 |
+
"num_tokens": 10211239.0,
|
| 601 |
+
"step": 2350
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"entropy": 0.1380857729911804,
|
| 605 |
+
"epoch": 0.7014979905005481,
|
| 606 |
+
"grad_norm": 0.32876694202423096,
|
| 607 |
+
"learning_rate": 0.00015800361518377186,
|
| 608 |
+
"loss": 0.14104883193969728,
|
| 609 |
+
"mean_token_accuracy": 0.9470359092950821,
|
| 610 |
+
"num_tokens": 10430005.0,
|
| 611 |
+
"step": 2400
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.7014979905005481,
|
| 615 |
+
"eval_entropy": 0.12756498554500484,
|
| 616 |
+
"eval_loss": 0.13620829582214355,
|
| 617 |
+
"eval_mean_token_accuracy": 0.9487966088925378,
|
| 618 |
+
"eval_num_tokens": 10430005.0,
|
| 619 |
+
"eval_runtime": 51.2201,
|
| 620 |
+
"eval_samples_per_second": 22.999,
|
| 621 |
+
"eval_steps_per_second": 5.759,
|
| 622 |
+
"step": 2400
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"entropy": 0.13768024053424596,
|
| 626 |
+
"epoch": 0.7161125319693095,
|
| 627 |
+
"grad_norm": 0.2406591773033142,
|
| 628 |
+
"learning_rate": 0.00015699939746937136,
|
| 629 |
+
"loss": 0.13825586318969726,
|
| 630 |
+
"mean_token_accuracy": 0.9467134064435959,
|
| 631 |
+
"num_tokens": 10646506.0,
|
| 632 |
+
"step": 2450
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"entropy": 0.13076995383948087,
|
| 636 |
+
"epoch": 0.7307270734380709,
|
| 637 |
+
"grad_norm": 0.3014162480831146,
|
| 638 |
+
"learning_rate": 0.0001559951797549709,
|
| 639 |
+
"loss": 0.1335157871246338,
|
| 640 |
+
"mean_token_accuracy": 0.9489013949036598,
|
| 641 |
+
"num_tokens": 10863443.0,
|
| 642 |
+
"step": 2500
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"entropy": 0.12944862838834525,
|
| 646 |
+
"epoch": 0.7453416149068323,
|
| 647 |
+
"grad_norm": 0.2716672718524933,
|
| 648 |
+
"learning_rate": 0.0001549909620405704,
|
| 649 |
+
"loss": 0.13044677734375,
|
| 650 |
+
"mean_token_accuracy": 0.9514863756299019,
|
| 651 |
+
"num_tokens": 11081801.0,
|
| 652 |
+
"step": 2550
|
| 653 |
+
},
|
| 654 |
+
{
|
| 655 |
+
"entropy": 0.13712387483566998,
|
| 656 |
+
"epoch": 0.7599561563755937,
|
| 657 |
+
"grad_norm": 0.2736733555793762,
|
| 658 |
+
"learning_rate": 0.00015398674432616993,
|
| 659 |
+
"loss": 0.13736340522766113,
|
| 660 |
+
"mean_token_accuracy": 0.9477300536632538,
|
| 661 |
+
"num_tokens": 11300525.0,
|
| 662 |
+
"step": 2600
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.7599561563755937,
|
| 666 |
+
"eval_entropy": 0.1276566566552146,
|
| 667 |
+
"eval_loss": 0.13106825947761536,
|
| 668 |
+
"eval_mean_token_accuracy": 0.9497191014936415,
|
| 669 |
+
"eval_num_tokens": 11300525.0,
|
| 670 |
+
"eval_runtime": 51.2416,
|
| 671 |
+
"eval_samples_per_second": 22.989,
|
| 672 |
+
"eval_steps_per_second": 5.757,
|
| 673 |
+
"step": 2600
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"entropy": 0.1264355194196105,
|
| 677 |
+
"epoch": 0.7745706978443552,
|
| 678 |
+
"grad_norm": 0.2751106321811676,
|
| 679 |
+
"learning_rate": 0.00015298252661176943,
|
| 680 |
+
"loss": 0.12769680976867676,
|
| 681 |
+
"mean_token_accuracy": 0.9521830633282662,
|
| 682 |
+
"num_tokens": 11518418.0,
|
| 683 |
+
"step": 2650
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"entropy": 0.13140305127948523,
|
| 687 |
+
"epoch": 0.7891852393131166,
|
| 688 |
+
"grad_norm": 0.28094226121902466,
|
| 689 |
+
"learning_rate": 0.00015197830889736896,
|
| 690 |
+
"loss": 0.1322249412536621,
|
| 691 |
+
"mean_token_accuracy": 0.9508191779255867,
|
| 692 |
+
"num_tokens": 11733682.0,
|
| 693 |
+
"step": 2700
|
| 694 |
+
},
|
| 695 |
+
{
|
| 696 |
+
"entropy": 0.1296489790827036,
|
| 697 |
+
"epoch": 0.803799780781878,
|
| 698 |
+
"grad_norm": 0.31960225105285645,
|
| 699 |
+
"learning_rate": 0.0001509740911829685,
|
| 700 |
+
"loss": 0.13073083877563477,
|
| 701 |
+
"mean_token_accuracy": 0.9499760115146637,
|
| 702 |
+
"num_tokens": 11950207.0,
|
| 703 |
+
"step": 2750
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"entropy": 0.1249570251069963,
|
| 707 |
+
"epoch": 0.8184143222506394,
|
| 708 |
+
"grad_norm": 0.2821206748485565,
|
| 709 |
+
"learning_rate": 0.000149969873468568,
|
| 710 |
+
"loss": 0.12826437950134278,
|
| 711 |
+
"mean_token_accuracy": 0.9511147084832191,
|
| 712 |
+
"num_tokens": 12166720.0,
|
| 713 |
+
"step": 2800
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.8184143222506394,
|
| 717 |
+
"eval_entropy": 0.12591243864873708,
|
| 718 |
+
"eval_loss": 0.13125115633010864,
|
| 719 |
+
"eval_mean_token_accuracy": 0.9507922431169931,
|
| 720 |
+
"eval_num_tokens": 12166720.0,
|
| 721 |
+
"eval_runtime": 51.205,
|
| 722 |
+
"eval_samples_per_second": 23.006,
|
| 723 |
+
"eval_steps_per_second": 5.761,
|
| 724 |
+
"step": 2800
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"entropy": 0.12767981396988035,
|
| 728 |
+
"epoch": 0.8330288637194008,
|
| 729 |
+
"grad_norm": 0.27759984135627747,
|
| 730 |
+
"learning_rate": 0.00014896565575416753,
|
| 731 |
+
"loss": 0.12868986129760743,
|
| 732 |
+
"mean_token_accuracy": 0.9509958475828171,
|
| 733 |
+
"num_tokens": 12382797.0,
|
| 734 |
+
"step": 2850
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"entropy": 0.1295067841745913,
|
| 738 |
+
"epoch": 0.8476434051881622,
|
| 739 |
+
"grad_norm": 0.23248647153377533,
|
| 740 |
+
"learning_rate": 0.00014796143803976703,
|
| 741 |
+
"loss": 0.13320878982543946,
|
| 742 |
+
"mean_token_accuracy": 0.9507122281193733,
|
| 743 |
+
"num_tokens": 12601076.0,
|
| 744 |
+
"step": 2900
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"entropy": 0.13079511255025864,
|
| 748 |
+
"epoch": 0.8622579466569237,
|
| 749 |
+
"grad_norm": 0.26745545864105225,
|
| 750 |
+
"learning_rate": 0.00014695722032536654,
|
| 751 |
+
"loss": 0.13269658088684083,
|
| 752 |
+
"mean_token_accuracy": 0.9507605847716332,
|
| 753 |
+
"num_tokens": 12816469.0,
|
| 754 |
+
"step": 2950
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"entropy": 0.1252501069754362,
|
| 758 |
+
"epoch": 0.8768724881256851,
|
| 759 |
+
"grad_norm": 0.2826436758041382,
|
| 760 |
+
"learning_rate": 0.00014595300261096607,
|
| 761 |
+
"loss": 0.12778679847717286,
|
| 762 |
+
"mean_token_accuracy": 0.9513759270310402,
|
| 763 |
+
"num_tokens": 13032433.0,
|
| 764 |
+
"step": 3000
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.8768724881256851,
|
| 768 |
+
"eval_entropy": 0.13185835869635565,
|
| 769 |
+
"eval_loss": 0.12732699513435364,
|
| 770 |
+
"eval_mean_token_accuracy": 0.9507037514347141,
|
| 771 |
+
"eval_num_tokens": 13032433.0,
|
| 772 |
+
"eval_runtime": 51.8861,
|
| 773 |
+
"eval_samples_per_second": 22.704,
|
| 774 |
+
"eval_steps_per_second": 5.686,
|
| 775 |
+
"step": 3000
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"entropy": 0.12554888619109988,
|
| 779 |
+
"epoch": 0.8914870295944465,
|
| 780 |
+
"grad_norm": 0.329559862613678,
|
| 781 |
+
"learning_rate": 0.00014494878489656557,
|
| 782 |
+
"loss": 0.12740483283996581,
|
| 783 |
+
"mean_token_accuracy": 0.9519036790728569,
|
| 784 |
+
"num_tokens": 13249653.0,
|
| 785 |
+
"step": 3050
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"entropy": 0.12633929964154958,
|
| 789 |
+
"epoch": 0.9061015710632079,
|
| 790 |
+
"grad_norm": 0.2994729280471802,
|
| 791 |
+
"learning_rate": 0.0001439445671821651,
|
| 792 |
+
"loss": 0.12879551887512208,
|
| 793 |
+
"mean_token_accuracy": 0.9523486059904098,
|
| 794 |
+
"num_tokens": 13466568.0,
|
| 795 |
+
"step": 3100
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"entropy": 0.1324564717710018,
|
| 799 |
+
"epoch": 0.9207161125319693,
|
| 800 |
+
"grad_norm": 0.2625322639942169,
|
| 801 |
+
"learning_rate": 0.00014294034946776463,
|
| 802 |
+
"loss": 0.13260350227355958,
|
| 803 |
+
"mean_token_accuracy": 0.950848217010498,
|
| 804 |
+
"num_tokens": 13684721.0,
|
| 805 |
+
"step": 3150
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"entropy": 0.1254732285439968,
|
| 809 |
+
"epoch": 0.9353306540007307,
|
| 810 |
+
"grad_norm": 0.3398250639438629,
|
| 811 |
+
"learning_rate": 0.00014193613175336414,
|
| 812 |
+
"loss": 0.12758193016052247,
|
| 813 |
+
"mean_token_accuracy": 0.9529688519239425,
|
| 814 |
+
"num_tokens": 13903515.0,
|
| 815 |
+
"step": 3200
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 0.9353306540007307,
|
| 819 |
+
"eval_entropy": 0.12104560298434759,
|
| 820 |
+
"eval_loss": 0.1254371851682663,
|
| 821 |
+
"eval_mean_token_accuracy": 0.9526172644000942,
|
| 822 |
+
"eval_num_tokens": 13903515.0,
|
| 823 |
+
"eval_runtime": 51.6986,
|
| 824 |
+
"eval_samples_per_second": 22.786,
|
| 825 |
+
"eval_steps_per_second": 5.706,
|
| 826 |
+
"step": 3200
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"entropy": 0.13165340654551982,
|
| 830 |
+
"epoch": 0.9499451954694922,
|
| 831 |
+
"grad_norm": 0.32547247409820557,
|
| 832 |
+
"learning_rate": 0.00014093191403896367,
|
| 833 |
+
"loss": 0.13361644744873047,
|
| 834 |
+
"mean_token_accuracy": 0.9494915246963501,
|
| 835 |
+
"num_tokens": 14122365.0,
|
| 836 |
+
"step": 3250
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"entropy": 0.12670642552897335,
|
| 840 |
+
"epoch": 0.9645597369382536,
|
| 841 |
+
"grad_norm": 0.43059042096138,
|
| 842 |
+
"learning_rate": 0.00013992769632456317,
|
| 843 |
+
"loss": 0.12734654426574707,
|
| 844 |
+
"mean_token_accuracy": 0.9540221604704857,
|
| 845 |
+
"num_tokens": 14341363.0,
|
| 846 |
+
"step": 3300
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"entropy": 0.12478693343698978,
|
| 850 |
+
"epoch": 0.979174278407015,
|
| 851 |
+
"grad_norm": 0.24067845940589905,
|
| 852 |
+
"learning_rate": 0.00013892347861016268,
|
| 853 |
+
"loss": 0.12505006790161133,
|
| 854 |
+
"mean_token_accuracy": 0.9537291714549064,
|
| 855 |
+
"num_tokens": 14558651.0,
|
| 856 |
+
"step": 3350
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"entropy": 0.12454102255403995,
|
| 860 |
+
"epoch": 0.9937888198757764,
|
| 861 |
+
"grad_norm": 0.27057918906211853,
|
| 862 |
+
"learning_rate": 0.0001379192608957622,
|
| 863 |
+
"loss": 0.12519509315490723,
|
| 864 |
+
"mean_token_accuracy": 0.9531420814990997,
|
| 865 |
+
"num_tokens": 14774887.0,
|
| 866 |
+
"step": 3400
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.9937888198757764,
|
| 870 |
+
"eval_entropy": 0.1218604107150587,
|
| 871 |
+
"eval_loss": 0.12476576864719391,
|
| 872 |
+
"eval_mean_token_accuracy": 0.9522026013519804,
|
| 873 |
+
"eval_num_tokens": 14774887.0,
|
| 874 |
+
"eval_runtime": 52.2754,
|
| 875 |
+
"eval_samples_per_second": 22.535,
|
| 876 |
+
"eval_steps_per_second": 5.643,
|
| 877 |
+
"step": 3400
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"entropy": 0.11712383136607063,
|
| 881 |
+
"epoch": 1.0081841432225065,
|
| 882 |
+
"grad_norm": 0.4069111943244934,
|
| 883 |
+
"learning_rate": 0.0001369150431813617,
|
| 884 |
+
"loss": 0.11227139472961425,
|
| 885 |
+
"mean_token_accuracy": 0.957756791320549,
|
| 886 |
+
"num_tokens": 14988925.0,
|
| 887 |
+
"step": 3450
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"entropy": 0.10819806138053537,
|
| 891 |
+
"epoch": 1.0227986846912678,
|
| 892 |
+
"grad_norm": 0.347784161567688,
|
| 893 |
+
"learning_rate": 0.00013591082546696124,
|
| 894 |
+
"loss": 0.10597694396972657,
|
| 895 |
+
"mean_token_accuracy": 0.9602300328016281,
|
| 896 |
+
"num_tokens": 15207529.0,
|
| 897 |
+
"step": 3500
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"entropy": 0.09677543576806784,
|
| 901 |
+
"epoch": 1.0374132261600293,
|
| 902 |
+
"grad_norm": 0.43760985136032104,
|
| 903 |
+
"learning_rate": 0.00013490660775256077,
|
| 904 |
+
"loss": 0.0967884635925293,
|
| 905 |
+
"mean_token_accuracy": 0.9631710237264633,
|
| 906 |
+
"num_tokens": 15424380.0,
|
| 907 |
+
"step": 3550
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"entropy": 0.09595708003267646,
|
| 911 |
+
"epoch": 1.0520277676287906,
|
| 912 |
+
"grad_norm": 0.3321962356567383,
|
| 913 |
+
"learning_rate": 0.00013390239003816028,
|
| 914 |
+
"loss": 0.09627270698547363,
|
| 915 |
+
"mean_token_accuracy": 0.9640694811940194,
|
| 916 |
+
"num_tokens": 15640971.0,
|
| 917 |
+
"step": 3600
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 1.0520277676287906,
|
| 921 |
+
"eval_entropy": 0.10276495573126664,
|
| 922 |
+
"eval_loss": 0.1241004690527916,
|
| 923 |
+
"eval_mean_token_accuracy": 0.9543731022689302,
|
| 924 |
+
"eval_num_tokens": 15640971.0,
|
| 925 |
+
"eval_runtime": 51.458,
|
| 926 |
+
"eval_samples_per_second": 22.892,
|
| 927 |
+
"eval_steps_per_second": 5.733,
|
| 928 |
+
"step": 3600
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"entropy": 0.09977266293019056,
|
| 932 |
+
"epoch": 1.066642309097552,
|
| 933 |
+
"grad_norm": 0.34104689955711365,
|
| 934 |
+
"learning_rate": 0.0001328981723237598,
|
| 935 |
+
"loss": 0.09645257949829102,
|
| 936 |
+
"mean_token_accuracy": 0.9637984240055084,
|
| 937 |
+
"num_tokens": 15858524.0,
|
| 938 |
+
"step": 3650
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"entropy": 0.097247479185462,
|
| 942 |
+
"epoch": 1.0812568505663136,
|
| 943 |
+
"grad_norm": 0.29214659333229065,
|
| 944 |
+
"learning_rate": 0.0001318939546093593,
|
| 945 |
+
"loss": 0.09814267158508301,
|
| 946 |
+
"mean_token_accuracy": 0.9640810504555702,
|
| 947 |
+
"num_tokens": 16076800.0,
|
| 948 |
+
"step": 3700
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"entropy": 0.10092583011835814,
|
| 952 |
+
"epoch": 1.0958713920350749,
|
| 953 |
+
"grad_norm": 0.27324506640434265,
|
| 954 |
+
"learning_rate": 0.00013088973689495882,
|
| 955 |
+
"loss": 0.09999562263488769,
|
| 956 |
+
"mean_token_accuracy": 0.9627196389436722,
|
| 957 |
+
"num_tokens": 16295096.0,
|
| 958 |
+
"step": 3750
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"entropy": 0.09513680284842849,
|
| 962 |
+
"epoch": 1.1104859335038364,
|
| 963 |
+
"grad_norm": 0.32636934518814087,
|
| 964 |
+
"learning_rate": 0.00012988551918055835,
|
| 965 |
+
"loss": 0.09587624549865723,
|
| 966 |
+
"mean_token_accuracy": 0.9640011402964592,
|
| 967 |
+
"num_tokens": 16511062.0,
|
| 968 |
+
"step": 3800
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.1104859335038364,
|
| 972 |
+
"eval_entropy": 0.09597434103488922,
|
| 973 |
+
"eval_loss": 0.12425629794597626,
|
| 974 |
+
"eval_mean_token_accuracy": 0.9549674963547011,
|
| 975 |
+
"eval_num_tokens": 16511062.0,
|
| 976 |
+
"eval_runtime": 51.6304,
|
| 977 |
+
"eval_samples_per_second": 22.816,
|
| 978 |
+
"eval_steps_per_second": 5.714,
|
| 979 |
+
"step": 3800
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"entropy": 0.09476003058254719,
|
| 983 |
+
"epoch": 1.1251004749725977,
|
| 984 |
+
"grad_norm": 0.28826117515563965,
|
| 985 |
+
"learning_rate": 0.00012888130146615788,
|
| 986 |
+
"loss": 0.09343017578125,
|
| 987 |
+
"mean_token_accuracy": 0.9631267023086548,
|
| 988 |
+
"num_tokens": 16728855.0,
|
| 989 |
+
"step": 3850
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"entropy": 0.09374131880700588,
|
| 993 |
+
"epoch": 1.1397150164413592,
|
| 994 |
+
"grad_norm": 0.3931892514228821,
|
| 995 |
+
"learning_rate": 0.00012787708375175738,
|
| 996 |
+
"loss": 0.09412909507751464,
|
| 997 |
+
"mean_token_accuracy": 0.9654038622975349,
|
| 998 |
+
"num_tokens": 16944362.0,
|
| 999 |
+
"step": 3900
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"entropy": 0.09230754490941763,
|
| 1003 |
+
"epoch": 1.1543295579101205,
|
| 1004 |
+
"grad_norm": 0.3274497389793396,
|
| 1005 |
+
"learning_rate": 0.00012687286603735691,
|
| 1006 |
+
"loss": 0.09314986228942872,
|
| 1007 |
+
"mean_token_accuracy": 0.9648771300911904,
|
| 1008 |
+
"num_tokens": 17161661.0,
|
| 1009 |
+
"step": 3950
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"entropy": 0.10461167734116315,
|
| 1013 |
+
"epoch": 1.168944099378882,
|
| 1014 |
+
"grad_norm": 0.3222503662109375,
|
| 1015 |
+
"learning_rate": 0.00012586864832295642,
|
| 1016 |
+
"loss": 0.10164319992065429,
|
| 1017 |
+
"mean_token_accuracy": 0.9613501909375191,
|
| 1018 |
+
"num_tokens": 17379999.0,
|
| 1019 |
+
"step": 4000
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 1.168944099378882,
|
| 1023 |
+
"eval_entropy": 0.09555989803146507,
|
| 1024 |
+
"eval_loss": 0.1204494908452034,
|
| 1025 |
+
"eval_mean_token_accuracy": 0.9567918892634117,
|
| 1026 |
+
"eval_num_tokens": 17379999.0,
|
| 1027 |
+
"eval_runtime": 51.2346,
|
| 1028 |
+
"eval_samples_per_second": 22.992,
|
| 1029 |
+
"eval_steps_per_second": 5.758,
|
| 1030 |
+
"step": 4000
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"entropy": 0.09500436037778855,
|
| 1034 |
+
"epoch": 1.1835586408476435,
|
| 1035 |
+
"grad_norm": 0.34208106994628906,
|
| 1036 |
+
"learning_rate": 0.00012486443060855595,
|
| 1037 |
+
"loss": 0.09681806564331055,
|
| 1038 |
+
"mean_token_accuracy": 0.9633791667222976,
|
| 1039 |
+
"num_tokens": 17597011.0,
|
| 1040 |
+
"step": 4050
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"entropy": 0.09596794040873646,
|
| 1044 |
+
"epoch": 1.1981731823164048,
|
| 1045 |
+
"grad_norm": 0.6134909987449646,
|
| 1046 |
+
"learning_rate": 0.00012386021289415545,
|
| 1047 |
+
"loss": 0.09627909660339355,
|
| 1048 |
+
"mean_token_accuracy": 0.9645186159014701,
|
| 1049 |
+
"num_tokens": 17814323.0,
|
| 1050 |
+
"step": 4100
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"entropy": 0.09296734174713492,
|
| 1054 |
+
"epoch": 1.2127877237851663,
|
| 1055 |
+
"grad_norm": 0.309161514043808,
|
| 1056 |
+
"learning_rate": 0.00012285599517975496,
|
| 1057 |
+
"loss": 0.09468943595886231,
|
| 1058 |
+
"mean_token_accuracy": 0.9646199548244476,
|
| 1059 |
+
"num_tokens": 18030765.0,
|
| 1060 |
+
"step": 4150
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"entropy": 0.09392764896154404,
|
| 1064 |
+
"epoch": 1.2274022652539276,
|
| 1065 |
+
"grad_norm": 0.29440510272979736,
|
| 1066 |
+
"learning_rate": 0.0001218517774653545,
|
| 1067 |
+
"loss": 0.09385616302490235,
|
| 1068 |
+
"mean_token_accuracy": 0.9648527312278747,
|
| 1069 |
+
"num_tokens": 18247789.0,
|
| 1070 |
+
"step": 4200
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 1.2274022652539276,
|
| 1074 |
+
"eval_entropy": 0.0961840924822678,
|
| 1075 |
+
"eval_loss": 0.12064161151647568,
|
| 1076 |
+
"eval_mean_token_accuracy": 0.9558913909782798,
|
| 1077 |
+
"eval_num_tokens": 18247789.0,
|
| 1078 |
+
"eval_runtime": 51.771,
|
| 1079 |
+
"eval_samples_per_second": 22.754,
|
| 1080 |
+
"eval_steps_per_second": 5.698,
|
| 1081 |
+
"step": 4200
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"entropy": 0.09542508151382208,
|
| 1085 |
+
"epoch": 1.242016806722689,
|
| 1086 |
+
"grad_norm": 0.40045246481895447,
|
| 1087 |
+
"learning_rate": 0.000120847559750954,
|
| 1088 |
+
"loss": 0.09424283981323242,
|
| 1089 |
+
"mean_token_accuracy": 0.9646497869491577,
|
| 1090 |
+
"num_tokens": 18465958.0,
|
| 1091 |
+
"step": 4250
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"entropy": 0.09011594654526561,
|
| 1095 |
+
"epoch": 1.2566313481914504,
|
| 1096 |
+
"grad_norm": 0.265855610370636,
|
| 1097 |
+
"learning_rate": 0.00011984334203655354,
|
| 1098 |
+
"loss": 0.090297269821167,
|
| 1099 |
+
"mean_token_accuracy": 0.9658442854881286,
|
| 1100 |
+
"num_tokens": 18682990.0,
|
| 1101 |
+
"step": 4300
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"entropy": 0.08969980413094163,
|
| 1105 |
+
"epoch": 1.271245889660212,
|
| 1106 |
+
"grad_norm": 0.40770867466926575,
|
| 1107 |
+
"learning_rate": 0.00011883912432215305,
|
| 1108 |
+
"loss": 0.09162886619567871,
|
| 1109 |
+
"mean_token_accuracy": 0.9658057001233101,
|
| 1110 |
+
"num_tokens": 18900419.0,
|
| 1111 |
+
"step": 4350
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"entropy": 0.09763234948739409,
|
| 1115 |
+
"epoch": 1.2858604311289734,
|
| 1116 |
+
"grad_norm": 0.3054230511188507,
|
| 1117 |
+
"learning_rate": 0.00011783490660775256,
|
| 1118 |
+
"loss": 0.0983086109161377,
|
| 1119 |
+
"mean_token_accuracy": 0.9633460545539856,
|
| 1120 |
+
"num_tokens": 19118904.0,
|
| 1121 |
+
"step": 4400
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 1.2858604311289734,
|
| 1125 |
+
"eval_entropy": 0.09841432020825855,
|
| 1126 |
+
"eval_loss": 0.1181686744093895,
|
| 1127 |
+
"eval_mean_token_accuracy": 0.9566149655034987,
|
| 1128 |
+
"eval_num_tokens": 19118904.0,
|
| 1129 |
+
"eval_runtime": 51.5747,
|
| 1130 |
+
"eval_samples_per_second": 22.841,
|
| 1131 |
+
"eval_steps_per_second": 5.72,
|
| 1132 |
+
"step": 4400
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"entropy": 0.09089874122291804,
|
| 1136 |
+
"epoch": 1.3004749725977347,
|
| 1137 |
+
"grad_norm": 0.36007627844810486,
|
| 1138 |
+
"learning_rate": 0.00011683068889335209,
|
| 1139 |
+
"loss": 0.08920018196105957,
|
| 1140 |
+
"mean_token_accuracy": 0.9661843276023865,
|
| 1141 |
+
"num_tokens": 19336274.0,
|
| 1142 |
+
"step": 4450
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"entropy": 0.0862886114232242,
|
| 1146 |
+
"epoch": 1.3150895140664962,
|
| 1147 |
+
"grad_norm": 0.33618125319480896,
|
| 1148 |
+
"learning_rate": 0.00011582647117895161,
|
| 1149 |
+
"loss": 0.08620091438293458,
|
| 1150 |
+
"mean_token_accuracy": 0.9683165431022644,
|
| 1151 |
+
"num_tokens": 19552793.0,
|
| 1152 |
+
"step": 4500
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"entropy": 0.0956742905639112,
|
| 1156 |
+
"epoch": 1.3297040555352575,
|
| 1157 |
+
"grad_norm": 0.3849150538444519,
|
| 1158 |
+
"learning_rate": 0.00011482225346455111,
|
| 1159 |
+
"loss": 0.09665518760681152,
|
| 1160 |
+
"mean_token_accuracy": 0.9633703279495239,
|
| 1161 |
+
"num_tokens": 19769345.0,
|
| 1162 |
+
"step": 4550
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"entropy": 0.09203850487247109,
|
| 1166 |
+
"epoch": 1.344318597004019,
|
| 1167 |
+
"grad_norm": 0.31663480401039124,
|
| 1168 |
+
"learning_rate": 0.00011381803575015064,
|
| 1169 |
+
"loss": 0.09010299682617187,
|
| 1170 |
+
"mean_token_accuracy": 0.9652843907475471,
|
| 1171 |
+
"num_tokens": 19987045.0,
|
| 1172 |
+
"step": 4600
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 1.344318597004019,
|
| 1176 |
+
"eval_entropy": 0.08627223369302386,
|
| 1177 |
+
"eval_loss": 0.12027477473020554,
|
| 1178 |
+
"eval_mean_token_accuracy": 0.9569575905799865,
|
| 1179 |
+
"eval_num_tokens": 19987045.0,
|
| 1180 |
+
"eval_runtime": 51.593,
|
| 1181 |
+
"eval_samples_per_second": 22.833,
|
| 1182 |
+
"eval_steps_per_second": 5.718,
|
| 1183 |
+
"step": 4600
|
| 1184 |
+
},
|
| 1185 |
+
{
|
| 1186 |
+
"entropy": 0.0923673815280199,
|
| 1187 |
+
"epoch": 1.3589331384727803,
|
| 1188 |
+
"grad_norm": 0.40604791045188904,
|
| 1189 |
+
"learning_rate": 0.00011281381803575015,
|
| 1190 |
+
"loss": 0.09391559600830078,
|
| 1191 |
+
"mean_token_accuracy": 0.9660267063975334,
|
| 1192 |
+
"num_tokens": 20204662.0,
|
| 1193 |
+
"step": 4650
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"entropy": 0.08946419894695282,
|
| 1197 |
+
"epoch": 1.3735476799415418,
|
| 1198 |
+
"grad_norm": 0.44695547223091125,
|
| 1199 |
+
"learning_rate": 0.00011180960032134968,
|
| 1200 |
+
"loss": 0.09293787002563476,
|
| 1201 |
+
"mean_token_accuracy": 0.9646873572468757,
|
| 1202 |
+
"num_tokens": 20423466.0,
|
| 1203 |
+
"step": 4700
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"entropy": 0.09025523517280817,
|
| 1207 |
+
"epoch": 1.3881622214103033,
|
| 1208 |
+
"grad_norm": 0.2976325750350952,
|
| 1209 |
+
"learning_rate": 0.0001108053826069492,
|
| 1210 |
+
"loss": 0.0911507797241211,
|
| 1211 |
+
"mean_token_accuracy": 0.9653226399421692,
|
| 1212 |
+
"num_tokens": 20641293.0,
|
| 1213 |
+
"step": 4750
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"entropy": 0.09211947876960039,
|
| 1217 |
+
"epoch": 1.4027767628790646,
|
| 1218 |
+
"grad_norm": 0.26411667466163635,
|
| 1219 |
+
"learning_rate": 0.0001098011648925487,
|
| 1220 |
+
"loss": 0.09012277603149414,
|
| 1221 |
+
"mean_token_accuracy": 0.9667563599348068,
|
| 1222 |
+
"num_tokens": 20859728.0,
|
| 1223 |
+
"step": 4800
|
| 1224 |
+
},
|
| 1225 |
+
{
|
| 1226 |
+
"epoch": 1.4027767628790646,
|
| 1227 |
+
"eval_entropy": 0.0889583650786998,
|
| 1228 |
+
"eval_loss": 0.11991087347269058,
|
| 1229 |
+
"eval_mean_token_accuracy": 0.9574164336010561,
|
| 1230 |
+
"eval_num_tokens": 20859728.0,
|
| 1231 |
+
"eval_runtime": 51.832,
|
| 1232 |
+
"eval_samples_per_second": 22.727,
|
| 1233 |
+
"eval_steps_per_second": 5.691,
|
| 1234 |
+
"step": 4800
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"entropy": 0.09142036261036993,
|
| 1238 |
+
"epoch": 1.4173913043478261,
|
| 1239 |
+
"grad_norm": 0.3087850511074066,
|
| 1240 |
+
"learning_rate": 0.00010879694717814823,
|
| 1241 |
+
"loss": 0.09248518943786621,
|
| 1242 |
+
"mean_token_accuracy": 0.9662388986349106,
|
| 1243 |
+
"num_tokens": 21077745.0,
|
| 1244 |
+
"step": 4850
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"entropy": 0.09305513446219266,
|
| 1248 |
+
"epoch": 1.4320058458165876,
|
| 1249 |
+
"grad_norm": 0.34897470474243164,
|
| 1250 |
+
"learning_rate": 0.00010779272946374775,
|
| 1251 |
+
"loss": 0.09240678787231445,
|
| 1252 |
+
"mean_token_accuracy": 0.9652760860323906,
|
| 1253 |
+
"num_tokens": 21295057.0,
|
| 1254 |
+
"step": 4900
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"entropy": 0.09016769832000136,
|
| 1258 |
+
"epoch": 1.446620387285349,
|
| 1259 |
+
"grad_norm": 0.32600948214530945,
|
| 1260 |
+
"learning_rate": 0.00010678851174934725,
|
| 1261 |
+
"loss": 0.0925068473815918,
|
| 1262 |
+
"mean_token_accuracy": 0.9654142627120018,
|
| 1263 |
+
"num_tokens": 21509590.0,
|
| 1264 |
+
"step": 4950
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"entropy": 0.08982307212427258,
|
| 1268 |
+
"epoch": 1.4612349287541102,
|
| 1269 |
+
"grad_norm": 0.3035278916358948,
|
| 1270 |
+
"learning_rate": 0.00010578429403494678,
|
| 1271 |
+
"loss": 0.09122074127197266,
|
| 1272 |
+
"mean_token_accuracy": 0.96599131077528,
|
| 1273 |
+
"num_tokens": 21727721.0,
|
| 1274 |
+
"step": 5000
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 1.4612349287541102,
|
| 1278 |
+
"eval_entropy": 0.09071531677145069,
|
| 1279 |
+
"eval_loss": 0.11822337657213211,
|
| 1280 |
+
"eval_mean_token_accuracy": 0.9580705240621405,
|
| 1281 |
+
"eval_num_tokens": 21727721.0,
|
| 1282 |
+
"eval_runtime": 51.3724,
|
| 1283 |
+
"eval_samples_per_second": 22.931,
|
| 1284 |
+
"eval_steps_per_second": 5.742,
|
| 1285 |
+
"step": 5000
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"entropy": 0.0927383072860539,
|
| 1289 |
+
"epoch": 1.4758494702228717,
|
| 1290 |
+
"grad_norm": 0.33800584077835083,
|
| 1291 |
+
"learning_rate": 0.0001047800763205463,
|
| 1292 |
+
"loss": 0.09140531539916992,
|
| 1293 |
+
"mean_token_accuracy": 0.9651796925067901,
|
| 1294 |
+
"num_tokens": 21945922.0,
|
| 1295 |
+
"step": 5050
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"entropy": 0.09029178963974119,
|
| 1299 |
+
"epoch": 1.4904640116916332,
|
| 1300 |
+
"grad_norm": 0.4534330368041992,
|
| 1301 |
+
"learning_rate": 0.00010377585860614583,
|
| 1302 |
+
"loss": 0.09171391487121582,
|
| 1303 |
+
"mean_token_accuracy": 0.966952161192894,
|
| 1304 |
+
"num_tokens": 22164989.0,
|
| 1305 |
+
"step": 5100
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"entropy": 0.0846068394370377,
|
| 1309 |
+
"epoch": 1.5050785531603945,
|
| 1310 |
+
"grad_norm": 0.2575761079788208,
|
| 1311 |
+
"learning_rate": 0.00010277164089174534,
|
| 1312 |
+
"loss": 0.08573863983154296,
|
| 1313 |
+
"mean_token_accuracy": 0.9693089470267295,
|
| 1314 |
+
"num_tokens": 22381112.0,
|
| 1315 |
+
"step": 5150
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"entropy": 0.08938812999054789,
|
| 1319 |
+
"epoch": 1.519693094629156,
|
| 1320 |
+
"grad_norm": 0.2744484543800354,
|
| 1321 |
+
"learning_rate": 0.00010176742317734484,
|
| 1322 |
+
"loss": 0.08850445747375488,
|
| 1323 |
+
"mean_token_accuracy": 0.9676422467827797,
|
| 1324 |
+
"num_tokens": 22597002.0,
|
| 1325 |
+
"step": 5200
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.519693094629156,
|
| 1329 |
+
"eval_entropy": 0.0930880088169696,
|
| 1330 |
+
"eval_loss": 0.11626386642456055,
|
| 1331 |
+
"eval_mean_token_accuracy": 0.9575319203279786,
|
| 1332 |
+
"eval_num_tokens": 22597002.0,
|
| 1333 |
+
"eval_runtime": 51.3354,
|
| 1334 |
+
"eval_samples_per_second": 22.947,
|
| 1335 |
+
"eval_steps_per_second": 5.747,
|
| 1336 |
+
"step": 5200
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"entropy": 0.09117060529068112,
|
| 1340 |
+
"epoch": 1.5343076360979175,
|
| 1341 |
+
"grad_norm": 0.38833215832710266,
|
| 1342 |
+
"learning_rate": 0.00010076320546294437,
|
| 1343 |
+
"loss": 0.09367633819580078,
|
| 1344 |
+
"mean_token_accuracy": 0.9644442161917687,
|
| 1345 |
+
"num_tokens": 22812799.0,
|
| 1346 |
+
"step": 5250
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"entropy": 0.08938059702515602,
|
| 1350 |
+
"epoch": 1.5489221775666788,
|
| 1351 |
+
"grad_norm": 0.3190608322620392,
|
| 1352 |
+
"learning_rate": 9.975898774854389e-05,
|
| 1353 |
+
"loss": 0.0879911231994629,
|
| 1354 |
+
"mean_token_accuracy": 0.96626698076725,
|
| 1355 |
+
"num_tokens": 23030016.0,
|
| 1356 |
+
"step": 5300
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"entropy": 0.08715751779265701,
|
| 1360 |
+
"epoch": 1.5635367190354401,
|
| 1361 |
+
"grad_norm": 0.30828171968460083,
|
| 1362 |
+
"learning_rate": 9.87547700341434e-05,
|
| 1363 |
+
"loss": 0.09027284622192383,
|
| 1364 |
+
"mean_token_accuracy": 0.9669826221466065,
|
| 1365 |
+
"num_tokens": 23248968.0,
|
| 1366 |
+
"step": 5350
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"entropy": 0.08828917387872934,
|
| 1370 |
+
"epoch": 1.5781512605042018,
|
| 1371 |
+
"grad_norm": 0.38410839438438416,
|
| 1372 |
+
"learning_rate": 9.775055231974292e-05,
|
| 1373 |
+
"loss": 0.08821515083312988,
|
| 1374 |
+
"mean_token_accuracy": 0.9665435764193535,
|
| 1375 |
+
"num_tokens": 23464933.0,
|
| 1376 |
+
"step": 5400
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 1.5781512605042018,
|
| 1380 |
+
"eval_entropy": 0.09166115953002946,
|
| 1381 |
+
"eval_loss": 0.11569290608167648,
|
| 1382 |
+
"eval_mean_token_accuracy": 0.9589312512995833,
|
| 1383 |
+
"eval_num_tokens": 23464933.0,
|
| 1384 |
+
"eval_runtime": 51.7313,
|
| 1385 |
+
"eval_samples_per_second": 22.771,
|
| 1386 |
+
"eval_steps_per_second": 5.703,
|
| 1387 |
+
"step": 5400
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"entropy": 0.08674560531042516,
|
| 1391 |
+
"epoch": 1.5927658019729631,
|
| 1392 |
+
"grad_norm": 0.2987992763519287,
|
| 1393 |
+
"learning_rate": 9.674633460534244e-05,
|
| 1394 |
+
"loss": 0.08608449935913086,
|
| 1395 |
+
"mean_token_accuracy": 0.9669453829526902,
|
| 1396 |
+
"num_tokens": 23683369.0,
|
| 1397 |
+
"step": 5450
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"entropy": 0.08784966805949807,
|
| 1401 |
+
"epoch": 1.6073803434417244,
|
| 1402 |
+
"grad_norm": 0.34559184312820435,
|
| 1403 |
+
"learning_rate": 9.574211689094196e-05,
|
| 1404 |
+
"loss": 0.08749322891235352,
|
| 1405 |
+
"mean_token_accuracy": 0.9666104671359063,
|
| 1406 |
+
"num_tokens": 23900573.0,
|
| 1407 |
+
"step": 5500
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"entropy": 0.08791282497346402,
|
| 1411 |
+
"epoch": 1.621994884910486,
|
| 1412 |
+
"grad_norm": 0.33519676327705383,
|
| 1413 |
+
"learning_rate": 9.473789917654148e-05,
|
| 1414 |
+
"loss": 0.08817813873291015,
|
| 1415 |
+
"mean_token_accuracy": 0.967940234541893,
|
| 1416 |
+
"num_tokens": 24117509.0,
|
| 1417 |
+
"step": 5550
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"entropy": 0.08461488509550691,
|
| 1421 |
+
"epoch": 1.6366094263792474,
|
| 1422 |
+
"grad_norm": 0.6715280413627625,
|
| 1423 |
+
"learning_rate": 9.3733681462141e-05,
|
| 1424 |
+
"loss": 0.08558347702026367,
|
| 1425 |
+
"mean_token_accuracy": 0.9682379782199859,
|
| 1426 |
+
"num_tokens": 24333931.0,
|
| 1427 |
+
"step": 5600
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 1.6366094263792474,
|
| 1431 |
+
"eval_entropy": 0.08775949452261804,
|
| 1432 |
+
"eval_loss": 0.11595170944929123,
|
| 1433 |
+
"eval_mean_token_accuracy": 0.9591477790121304,
|
| 1434 |
+
"eval_num_tokens": 24333931.0,
|
| 1435 |
+
"eval_runtime": 51.6207,
|
| 1436 |
+
"eval_samples_per_second": 22.82,
|
| 1437 |
+
"eval_steps_per_second": 5.715,
|
| 1438 |
+
"step": 5600
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"entropy": 0.08632316379807889,
|
| 1442 |
+
"epoch": 1.6512239678480087,
|
| 1443 |
+
"grad_norm": 0.426924467086792,
|
| 1444 |
+
"learning_rate": 9.272946374774053e-05,
|
| 1445 |
+
"loss": 0.08816647529602051,
|
| 1446 |
+
"mean_token_accuracy": 0.9676574739813805,
|
| 1447 |
+
"num_tokens": 24551230.0,
|
| 1448 |
+
"step": 5650
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"entropy": 0.08664186015725135,
|
| 1452 |
+
"epoch": 1.66583850931677,
|
| 1453 |
+
"grad_norm": 0.2743942439556122,
|
| 1454 |
+
"learning_rate": 9.172524603334003e-05,
|
| 1455 |
+
"loss": 0.08955062866210937,
|
| 1456 |
+
"mean_token_accuracy": 0.9671838331222534,
|
| 1457 |
+
"num_tokens": 24768343.0,
|
| 1458 |
+
"step": 5700
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"entropy": 0.0877827282436192,
|
| 1462 |
+
"epoch": 1.6804530507855318,
|
| 1463 |
+
"grad_norm": 0.4056563675403595,
|
| 1464 |
+
"learning_rate": 9.072102831893955e-05,
|
| 1465 |
+
"loss": 0.08868096351623535,
|
| 1466 |
+
"mean_token_accuracy": 0.9664111837744713,
|
| 1467 |
+
"num_tokens": 24984999.0,
|
| 1468 |
+
"step": 5750
|
| 1469 |
+
},
|
| 1470 |
+
{
|
| 1471 |
+
"entropy": 0.08376523592509329,
|
| 1472 |
+
"epoch": 1.695067592254293,
|
| 1473 |
+
"grad_norm": 0.43965309858322144,
|
| 1474 |
+
"learning_rate": 8.971681060453906e-05,
|
| 1475 |
+
"loss": 0.08181517601013183,
|
| 1476 |
+
"mean_token_accuracy": 0.9681954535841942,
|
| 1477 |
+
"num_tokens": 25202890.0,
|
| 1478 |
+
"step": 5800
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 1.695067592254293,
|
| 1482 |
+
"eval_entropy": 0.08046806757101568,
|
| 1483 |
+
"eval_loss": 0.11467474699020386,
|
| 1484 |
+
"eval_mean_token_accuracy": 0.959945414430004,
|
| 1485 |
+
"eval_num_tokens": 25202890.0,
|
| 1486 |
+
"eval_runtime": 51.7478,
|
| 1487 |
+
"eval_samples_per_second": 22.764,
|
| 1488 |
+
"eval_steps_per_second": 5.701,
|
| 1489 |
+
"step": 5800
|
| 1490 |
+
},
|
| 1491 |
+
{
|
| 1492 |
+
"entropy": 0.08683971673250199,
|
| 1493 |
+
"epoch": 1.7096821337230543,
|
| 1494 |
+
"grad_norm": 0.368268758058548,
|
| 1495 |
+
"learning_rate": 8.87125928901386e-05,
|
| 1496 |
+
"loss": 0.09294096946716308,
|
| 1497 |
+
"mean_token_accuracy": 0.9656963437795639,
|
| 1498 |
+
"num_tokens": 25419806.0,
|
| 1499 |
+
"step": 5850
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"entropy": 0.08786554461345077,
|
| 1503 |
+
"epoch": 1.7242966751918158,
|
| 1504 |
+
"grad_norm": 0.2730977237224579,
|
| 1505 |
+
"learning_rate": 8.77083751757381e-05,
|
| 1506 |
+
"loss": 0.0859497356414795,
|
| 1507 |
+
"mean_token_accuracy": 0.9679459264874458,
|
| 1508 |
+
"num_tokens": 25636607.0,
|
| 1509 |
+
"step": 5900
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"entropy": 0.0849998215213418,
|
| 1513 |
+
"epoch": 1.7389112166605774,
|
| 1514 |
+
"grad_norm": 0.38372334837913513,
|
| 1515 |
+
"learning_rate": 8.670415746133762e-05,
|
| 1516 |
+
"loss": 0.08449971199035644,
|
| 1517 |
+
"mean_token_accuracy": 0.9687798276543618,
|
| 1518 |
+
"num_tokens": 25853046.0,
|
| 1519 |
+
"step": 5950
|
| 1520 |
+
},
|
| 1521 |
+
{
|
| 1522 |
+
"entropy": 0.08693754209205508,
|
| 1523 |
+
"epoch": 1.7535257581293386,
|
| 1524 |
+
"grad_norm": 0.2625570297241211,
|
| 1525 |
+
"learning_rate": 8.569993974693714e-05,
|
| 1526 |
+
"loss": 0.0882598876953125,
|
| 1527 |
+
"mean_token_accuracy": 0.9681905370950699,
|
| 1528 |
+
"num_tokens": 26070424.0,
|
| 1529 |
+
"step": 6000
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 1.7535257581293386,
|
| 1533 |
+
"eval_entropy": 0.0895412382306689,
|
| 1534 |
+
"eval_loss": 0.11259053647518158,
|
| 1535 |
+
"eval_mean_token_accuracy": 0.959880271402456,
|
| 1536 |
+
"eval_num_tokens": 26070424.0,
|
| 1537 |
+
"eval_runtime": 51.2473,
|
| 1538 |
+
"eval_samples_per_second": 22.987,
|
| 1539 |
+
"eval_steps_per_second": 5.756,
|
| 1540 |
+
"step": 6000
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"entropy": 0.08510855933651328,
|
| 1544 |
+
"epoch": 1.7681402995981002,
|
| 1545 |
+
"grad_norm": 0.3207480013370514,
|
| 1546 |
+
"learning_rate": 8.469572203253667e-05,
|
| 1547 |
+
"loss": 0.0856710433959961,
|
| 1548 |
+
"mean_token_accuracy": 0.9684762534499168,
|
| 1549 |
+
"num_tokens": 26286207.0,
|
| 1550 |
+
"step": 6050
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"entropy": 0.08411331648007035,
|
| 1554 |
+
"epoch": 1.7827548410668617,
|
| 1555 |
+
"grad_norm": 0.353188693523407,
|
| 1556 |
+
"learning_rate": 8.369150431813617e-05,
|
| 1557 |
+
"loss": 0.0873355770111084,
|
| 1558 |
+
"mean_token_accuracy": 0.9668887314200402,
|
| 1559 |
+
"num_tokens": 26503987.0,
|
| 1560 |
+
"step": 6100
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"entropy": 0.08718796769157051,
|
| 1564 |
+
"epoch": 1.797369382535623,
|
| 1565 |
+
"grad_norm": 0.4751955270767212,
|
| 1566 |
+
"learning_rate": 8.268728660373569e-05,
|
| 1567 |
+
"loss": 0.08717143058776855,
|
| 1568 |
+
"mean_token_accuracy": 0.9682504817843437,
|
| 1569 |
+
"num_tokens": 26722694.0,
|
| 1570 |
+
"step": 6150
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"entropy": 0.08594496028497815,
|
| 1574 |
+
"epoch": 1.8119839240043842,
|
| 1575 |
+
"grad_norm": 0.3431340157985687,
|
| 1576 |
+
"learning_rate": 8.168306888933522e-05,
|
| 1577 |
+
"loss": 0.08702672004699707,
|
| 1578 |
+
"mean_token_accuracy": 0.9675016805529595,
|
| 1579 |
+
"num_tokens": 26940760.0,
|
| 1580 |
+
"step": 6200
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 1.8119839240043842,
|
| 1584 |
+
"eval_entropy": 0.09019251664058636,
|
| 1585 |
+
"eval_loss": 0.11118555068969727,
|
| 1586 |
+
"eval_mean_token_accuracy": 0.9598824753599652,
|
| 1587 |
+
"eval_num_tokens": 26940760.0,
|
| 1588 |
+
"eval_runtime": 51.3016,
|
| 1589 |
+
"eval_samples_per_second": 22.962,
|
| 1590 |
+
"eval_steps_per_second": 5.75,
|
| 1591 |
+
"step": 6200
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"entropy": 0.0879617197997868,
|
| 1595 |
+
"epoch": 1.8265984654731457,
|
| 1596 |
+
"grad_norm": 0.3889070153236389,
|
| 1597 |
+
"learning_rate": 8.067885117493474e-05,
|
| 1598 |
+
"loss": 0.08638958930969239,
|
| 1599 |
+
"mean_token_accuracy": 0.9678712266683579,
|
| 1600 |
+
"num_tokens": 27156798.0,
|
| 1601 |
+
"step": 6250
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"entropy": 0.08402345498092473,
|
| 1605 |
+
"epoch": 1.8412130069419073,
|
| 1606 |
+
"grad_norm": 0.3556425869464874,
|
| 1607 |
+
"learning_rate": 7.967463346053424e-05,
|
| 1608 |
+
"loss": 0.08525685310363769,
|
| 1609 |
+
"mean_token_accuracy": 0.9692794454097747,
|
| 1610 |
+
"num_tokens": 27374733.0,
|
| 1611 |
+
"step": 6300
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"entropy": 0.08294867504388094,
|
| 1615 |
+
"epoch": 1.8558275484106685,
|
| 1616 |
+
"grad_norm": 0.31357839703559875,
|
| 1617 |
+
"learning_rate": 7.867041574613376e-05,
|
| 1618 |
+
"loss": 0.08585151672363281,
|
| 1619 |
+
"mean_token_accuracy": 0.968332688510418,
|
| 1620 |
+
"num_tokens": 27592971.0,
|
| 1621 |
+
"step": 6350
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"entropy": 0.08433418484404683,
|
| 1625 |
+
"epoch": 1.87044208987943,
|
| 1626 |
+
"grad_norm": 0.376997172832489,
|
| 1627 |
+
"learning_rate": 7.766619803173329e-05,
|
| 1628 |
+
"loss": 0.08210147857666016,
|
| 1629 |
+
"mean_token_accuracy": 0.9694547387957573,
|
| 1630 |
+
"num_tokens": 27809346.0,
|
| 1631 |
+
"step": 6400
|
| 1632 |
+
},
|
| 1633 |
+
{
|
| 1634 |
+
"epoch": 1.87044208987943,
|
| 1635 |
+
"eval_entropy": 0.08374374596623041,
|
| 1636 |
+
"eval_loss": 0.11078612506389618,
|
| 1637 |
+
"eval_mean_token_accuracy": 0.9607059201951754,
|
| 1638 |
+
"eval_num_tokens": 27809346.0,
|
| 1639 |
+
"eval_runtime": 51.3823,
|
| 1640 |
+
"eval_samples_per_second": 22.926,
|
| 1641 |
+
"eval_steps_per_second": 5.741,
|
| 1642 |
+
"step": 6400
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"entropy": 0.08094374425709247,
|
| 1646 |
+
"epoch": 1.8850566313481916,
|
| 1647 |
+
"grad_norm": 0.35051146149635315,
|
| 1648 |
+
"learning_rate": 7.666198031733281e-05,
|
| 1649 |
+
"loss": 0.08127692222595215,
|
| 1650 |
+
"mean_token_accuracy": 0.9706088775396347,
|
| 1651 |
+
"num_tokens": 28025421.0,
|
| 1652 |
+
"step": 6450
|
| 1653 |
+
},
|
| 1654 |
+
{
|
| 1655 |
+
"entropy": 0.08181249966844917,
|
| 1656 |
+
"epoch": 1.8996711728169529,
|
| 1657 |
+
"grad_norm": 0.361937016248703,
|
| 1658 |
+
"learning_rate": 7.565776260293231e-05,
|
| 1659 |
+
"loss": 0.08228342056274414,
|
| 1660 |
+
"mean_token_accuracy": 0.9692467513680458,
|
| 1661 |
+
"num_tokens": 28241427.0,
|
| 1662 |
+
"step": 6500
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"entropy": 0.0853189903870225,
|
| 1666 |
+
"epoch": 1.9142857142857141,
|
| 1667 |
+
"grad_norm": 0.4127114713191986,
|
| 1668 |
+
"learning_rate": 7.465354488853183e-05,
|
| 1669 |
+
"loss": 0.08898340225219727,
|
| 1670 |
+
"mean_token_accuracy": 0.9675960612297058,
|
| 1671 |
+
"num_tokens": 28459034.0,
|
| 1672 |
+
"step": 6550
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"entropy": 0.08330891110002994,
|
| 1676 |
+
"epoch": 1.9289002557544757,
|
| 1677 |
+
"grad_norm": 0.31873515248298645,
|
| 1678 |
+
"learning_rate": 7.364932717413136e-05,
|
| 1679 |
+
"loss": 0.0857757568359375,
|
| 1680 |
+
"mean_token_accuracy": 0.9681384879350662,
|
| 1681 |
+
"num_tokens": 28676875.0,
|
| 1682 |
+
"step": 6600
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 1.9289002557544757,
|
| 1686 |
+
"eval_entropy": 0.08660914314500356,
|
| 1687 |
+
"eval_loss": 0.10761687159538269,
|
| 1688 |
+
"eval_mean_token_accuracy": 0.9621669098482294,
|
| 1689 |
+
"eval_num_tokens": 28676875.0,
|
| 1690 |
+
"eval_runtime": 51.5275,
|
| 1691 |
+
"eval_samples_per_second": 22.862,
|
| 1692 |
+
"eval_steps_per_second": 5.725,
|
| 1693 |
+
"step": 6600
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"entropy": 0.08828531488776208,
|
| 1697 |
+
"epoch": 1.9435147972232372,
|
| 1698 |
+
"grad_norm": 0.3040439486503601,
|
| 1699 |
+
"learning_rate": 7.264510945973088e-05,
|
| 1700 |
+
"loss": 0.08551553726196288,
|
| 1701 |
+
"mean_token_accuracy": 0.9684215313196183,
|
| 1702 |
+
"num_tokens": 28894213.0,
|
| 1703 |
+
"step": 6650
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"entropy": 0.08436274753883481,
|
| 1707 |
+
"epoch": 1.9581293386919985,
|
| 1708 |
+
"grad_norm": 0.3856823742389679,
|
| 1709 |
+
"learning_rate": 7.16408917453304e-05,
|
| 1710 |
+
"loss": 0.08532014846801758,
|
| 1711 |
+
"mean_token_accuracy": 0.9674273264408112,
|
| 1712 |
+
"num_tokens": 29111328.0,
|
| 1713 |
+
"step": 6700
|
| 1714 |
+
},
|
| 1715 |
+
{
|
| 1716 |
+
"entropy": 0.08388302940875292,
|
| 1717 |
+
"epoch": 1.97274388016076,
|
| 1718 |
+
"grad_norm": 0.3501090407371521,
|
| 1719 |
+
"learning_rate": 7.06366740309299e-05,
|
| 1720 |
+
"loss": 0.08266093254089356,
|
| 1721 |
+
"mean_token_accuracy": 0.9687711641192436,
|
| 1722 |
+
"num_tokens": 29328355.0,
|
| 1723 |
+
"step": 6750
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"entropy": 0.08621805243194103,
|
| 1727 |
+
"epoch": 1.9873584216295215,
|
| 1728 |
+
"grad_norm": 0.3734675347805023,
|
| 1729 |
+
"learning_rate": 6.963245631652943e-05,
|
| 1730 |
+
"loss": 0.08826584815979004,
|
| 1731 |
+
"mean_token_accuracy": 0.9675416949391366,
|
| 1732 |
+
"num_tokens": 29546388.0,
|
| 1733 |
+
"step": 6800
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 1.9873584216295215,
|
| 1737 |
+
"eval_entropy": 0.08420832851680658,
|
| 1738 |
+
"eval_loss": 0.10720802843570709,
|
| 1739 |
+
"eval_mean_token_accuracy": 0.9616350175970692,
|
| 1740 |
+
"eval_num_tokens": 29546388.0,
|
| 1741 |
+
"eval_runtime": 51.2571,
|
| 1742 |
+
"eval_samples_per_second": 22.982,
|
| 1743 |
+
"eval_steps_per_second": 5.755,
|
| 1744 |
+
"step": 6800
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"entropy": 0.08057132030501583,
|
| 1748 |
+
"epoch": 2.0017537449762512,
|
| 1749 |
+
"grad_norm": 0.25122901797294617,
|
| 1750 |
+
"learning_rate": 6.862823860212895e-05,
|
| 1751 |
+
"loss": 0.08081165313720703,
|
| 1752 |
+
"mean_token_accuracy": 0.9702659373356001,
|
| 1753 |
+
"num_tokens": 29759753.0,
|
| 1754 |
+
"step": 6850
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"entropy": 0.0640625470969826,
|
| 1758 |
+
"epoch": 2.016368286445013,
|
| 1759 |
+
"grad_norm": 0.2220752090215683,
|
| 1760 |
+
"learning_rate": 6.762402088772847e-05,
|
| 1761 |
+
"loss": 0.053548240661621095,
|
| 1762 |
+
"mean_token_accuracy": 0.9798859319090844,
|
| 1763 |
+
"num_tokens": 29977507.0,
|
| 1764 |
+
"step": 6900
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"entropy": 0.05466210062615574,
|
| 1768 |
+
"epoch": 2.0309828279137743,
|
| 1769 |
+
"grad_norm": 0.3461098074913025,
|
| 1770 |
+
"learning_rate": 6.661980317332798e-05,
|
| 1771 |
+
"loss": 0.049960670471191404,
|
| 1772 |
+
"mean_token_accuracy": 0.9817683389782905,
|
| 1773 |
+
"num_tokens": 30194924.0,
|
| 1774 |
+
"step": 6950
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"entropy": 0.05153279748279602,
|
| 1778 |
+
"epoch": 2.0455973693825356,
|
| 1779 |
+
"grad_norm": 0.4181481897830963,
|
| 1780 |
+
"learning_rate": 6.56155854589275e-05,
|
| 1781 |
+
"loss": 0.05096799373626709,
|
| 1782 |
+
"mean_token_accuracy": 0.9809333425760269,
|
| 1783 |
+
"num_tokens": 30411745.0,
|
| 1784 |
+
"step": 7000
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 2.0455973693825356,
|
| 1788 |
+
"eval_entropy": 0.05856578930486309,
|
| 1789 |
+
"eval_loss": 0.128422349691391,
|
| 1790 |
+
"eval_mean_token_accuracy": 0.9615406414209786,
|
| 1791 |
+
"eval_num_tokens": 30411745.0,
|
| 1792 |
+
"eval_runtime": 51.3762,
|
| 1793 |
+
"eval_samples_per_second": 22.929,
|
| 1794 |
+
"eval_steps_per_second": 5.742,
|
| 1795 |
+
"step": 7000
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"entropy": 0.048330898359417916,
|
| 1799 |
+
"epoch": 2.060211910851297,
|
| 1800 |
+
"grad_norm": 0.2673610746860504,
|
| 1801 |
+
"learning_rate": 6.461136774452702e-05,
|
| 1802 |
+
"loss": 0.04616006374359131,
|
| 1803 |
+
"mean_token_accuracy": 0.9836411371827125,
|
| 1804 |
+
"num_tokens": 30627471.0,
|
| 1805 |
+
"step": 7050
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"entropy": 0.05090990305878222,
|
| 1809 |
+
"epoch": 2.0748264523200586,
|
| 1810 |
+
"grad_norm": 0.37639951705932617,
|
| 1811 |
+
"learning_rate": 6.360715003012654e-05,
|
| 1812 |
+
"loss": 0.04985607624053955,
|
| 1813 |
+
"mean_token_accuracy": 0.9822920462489129,
|
| 1814 |
+
"num_tokens": 30845376.0,
|
| 1815 |
+
"step": 7100
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"entropy": 0.049039122625254096,
|
| 1819 |
+
"epoch": 2.08944099378882,
|
| 1820 |
+
"grad_norm": 0.28891128301620483,
|
| 1821 |
+
"learning_rate": 6.260293231572605e-05,
|
| 1822 |
+
"loss": 0.04553122997283936,
|
| 1823 |
+
"mean_token_accuracy": 0.9838429951667785,
|
| 1824 |
+
"num_tokens": 31062649.0,
|
| 1825 |
+
"step": 7150
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"entropy": 0.04845839339308441,
|
| 1829 |
+
"epoch": 2.104055535257581,
|
| 1830 |
+
"grad_norm": 0.37833932042121887,
|
| 1831 |
+
"learning_rate": 6.159871460132557e-05,
|
| 1832 |
+
"loss": 0.04832898139953613,
|
| 1833 |
+
"mean_token_accuracy": 0.9824651059508324,
|
| 1834 |
+
"num_tokens": 31279048.0,
|
| 1835 |
+
"step": 7200
|
| 1836 |
+
},
|
| 1837 |
+
{
|
| 1838 |
+
"epoch": 2.104055535257581,
|
| 1839 |
+
"eval_entropy": 0.05895731328844519,
|
| 1840 |
+
"eval_loss": 0.1306932121515274,
|
| 1841 |
+
"eval_mean_token_accuracy": 0.9612127061617577,
|
| 1842 |
+
"eval_num_tokens": 31279048.0,
|
| 1843 |
+
"eval_runtime": 51.5427,
|
| 1844 |
+
"eval_samples_per_second": 22.855,
|
| 1845 |
+
"eval_steps_per_second": 5.723,
|
| 1846 |
+
"step": 7200
|
| 1847 |
+
},
|
| 1848 |
+
{
|
| 1849 |
+
"entropy": 0.05211819554679096,
|
| 1850 |
+
"epoch": 2.118670076726343,
|
| 1851 |
+
"grad_norm": 0.36787164211273193,
|
| 1852 |
+
"learning_rate": 6.059449688692509e-05,
|
| 1853 |
+
"loss": 0.0502163028717041,
|
| 1854 |
+
"mean_token_accuracy": 0.9812204048037529,
|
| 1855 |
+
"num_tokens": 31496006.0,
|
| 1856 |
+
"step": 7250
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"entropy": 0.047373109618201854,
|
| 1860 |
+
"epoch": 2.133284618195104,
|
| 1861 |
+
"grad_norm": 0.3374268114566803,
|
| 1862 |
+
"learning_rate": 5.959027917252461e-05,
|
| 1863 |
+
"loss": 0.04724916934967041,
|
| 1864 |
+
"mean_token_accuracy": 0.9817435780167579,
|
| 1865 |
+
"num_tokens": 31713707.0,
|
| 1866 |
+
"step": 7300
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"entropy": 0.0479678925126791,
|
| 1870 |
+
"epoch": 2.1478991596638655,
|
| 1871 |
+
"grad_norm": 0.21644751727581024,
|
| 1872 |
+
"learning_rate": 5.858606145812412e-05,
|
| 1873 |
+
"loss": 0.046577134132385255,
|
| 1874 |
+
"mean_token_accuracy": 0.9833640906214715,
|
| 1875 |
+
"num_tokens": 31931335.0,
|
| 1876 |
+
"step": 7350
|
| 1877 |
+
},
|
| 1878 |
+
{
|
| 1879 |
+
"entropy": 0.04867122542113066,
|
| 1880 |
+
"epoch": 2.162513701132627,
|
| 1881 |
+
"grad_norm": 0.4094817340373993,
|
| 1882 |
+
"learning_rate": 5.758184374372364e-05,
|
| 1883 |
+
"loss": 0.047378678321838376,
|
| 1884 |
+
"mean_token_accuracy": 0.9824044448137284,
|
| 1885 |
+
"num_tokens": 32149008.0,
|
| 1886 |
+
"step": 7400
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 2.162513701132627,
|
| 1890 |
+
"eval_entropy": 0.058307369441677956,
|
| 1891 |
+
"eval_loss": 0.1280326098203659,
|
| 1892 |
+
"eval_mean_token_accuracy": 0.9618316363480132,
|
| 1893 |
+
"eval_num_tokens": 32149008.0,
|
| 1894 |
+
"eval_runtime": 51.2281,
|
| 1895 |
+
"eval_samples_per_second": 22.995,
|
| 1896 |
+
"eval_steps_per_second": 5.759,
|
| 1897 |
+
"step": 7400
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"entropy": 0.04640938185621053,
|
| 1901 |
+
"epoch": 2.1771282426013885,
|
| 1902 |
+
"grad_norm": 0.7977092266082764,
|
| 1903 |
+
"learning_rate": 5.657762602932316e-05,
|
| 1904 |
+
"loss": 0.045011258125305174,
|
| 1905 |
+
"mean_token_accuracy": 0.9833449149131774,
|
| 1906 |
+
"num_tokens": 32365330.0,
|
| 1907 |
+
"step": 7450
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"entropy": 0.046531920516863466,
|
| 1911 |
+
"epoch": 2.1917427840701498,
|
| 1912 |
+
"grad_norm": 0.4321069121360779,
|
| 1913 |
+
"learning_rate": 5.557340831492268e-05,
|
| 1914 |
+
"loss": 0.046553001403808594,
|
| 1915 |
+
"mean_token_accuracy": 0.9836733216047286,
|
| 1916 |
+
"num_tokens": 32582571.0,
|
| 1917 |
+
"step": 7500
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"entropy": 0.04584709240123629,
|
| 1921 |
+
"epoch": 2.206357325538911,
|
| 1922 |
+
"grad_norm": 0.39226505160331726,
|
| 1923 |
+
"learning_rate": 5.4569190600522194e-05,
|
| 1924 |
+
"loss": 0.04660297393798828,
|
| 1925 |
+
"mean_token_accuracy": 0.9831815731525421,
|
| 1926 |
+
"num_tokens": 32798927.0,
|
| 1927 |
+
"step": 7550
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"entropy": 0.051464518043212594,
|
| 1931 |
+
"epoch": 2.220971867007673,
|
| 1932 |
+
"grad_norm": 0.5300610661506653,
|
| 1933 |
+
"learning_rate": 5.356497288612171e-05,
|
| 1934 |
+
"loss": 0.049462013244628907,
|
| 1935 |
+
"mean_token_accuracy": 0.9828921148180961,
|
| 1936 |
+
"num_tokens": 33017400.0,
|
| 1937 |
+
"step": 7600
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 2.220971867007673,
|
| 1941 |
+
"eval_entropy": 0.054104896524321225,
|
| 1942 |
+
"eval_loss": 0.1346907615661621,
|
| 1943 |
+
"eval_mean_token_accuracy": 0.9608211775957528,
|
| 1944 |
+
"eval_num_tokens": 33017400.0,
|
| 1945 |
+
"eval_runtime": 51.5259,
|
| 1946 |
+
"eval_samples_per_second": 22.862,
|
| 1947 |
+
"eval_steps_per_second": 5.725,
|
| 1948 |
+
"step": 7600
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"entropy": 0.044908167165704074,
|
| 1952 |
+
"epoch": 2.235586408476434,
|
| 1953 |
+
"grad_norm": 0.5431990623474121,
|
| 1954 |
+
"learning_rate": 5.256075517172123e-05,
|
| 1955 |
+
"loss": 0.0420997953414917,
|
| 1956 |
+
"mean_token_accuracy": 0.9843368935585022,
|
| 1957 |
+
"num_tokens": 33233710.0,
|
| 1958 |
+
"step": 7650
|
| 1959 |
+
},
|
| 1960 |
+
{
|
| 1961 |
+
"entropy": 0.047161051160655915,
|
| 1962 |
+
"epoch": 2.2502009499451954,
|
| 1963 |
+
"grad_norm": 0.3964262306690216,
|
| 1964 |
+
"learning_rate": 5.1556537457320754e-05,
|
| 1965 |
+
"loss": 0.047385578155517576,
|
| 1966 |
+
"mean_token_accuracy": 0.9828732433915138,
|
| 1967 |
+
"num_tokens": 33451237.0,
|
| 1968 |
+
"step": 7700
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"entropy": 0.04588393676560372,
|
| 1972 |
+
"epoch": 2.264815491413957,
|
| 1973 |
+
"grad_norm": 0.3231643736362457,
|
| 1974 |
+
"learning_rate": 5.0552319742920264e-05,
|
| 1975 |
+
"loss": 0.04522655963897705,
|
| 1976 |
+
"mean_token_accuracy": 0.9838127914071083,
|
| 1977 |
+
"num_tokens": 33667691.0,
|
| 1978 |
+
"step": 7750
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"entropy": 0.04474953535944223,
|
| 1982 |
+
"epoch": 2.2794300328827184,
|
| 1983 |
+
"grad_norm": 0.3903498947620392,
|
| 1984 |
+
"learning_rate": 4.954810202851978e-05,
|
| 1985 |
+
"loss": 0.04430769443511963,
|
| 1986 |
+
"mean_token_accuracy": 0.9832708036899567,
|
| 1987 |
+
"num_tokens": 33885773.0,
|
| 1988 |
+
"step": 7800
|
| 1989 |
+
},
|
| 1990 |
+
{
|
| 1991 |
+
"epoch": 2.2794300328827184,
|
| 1992 |
+
"eval_entropy": 0.05170546539948653,
|
| 1993 |
+
"eval_loss": 0.13600485026836395,
|
| 1994 |
+
"eval_mean_token_accuracy": 0.9620883953773369,
|
| 1995 |
+
"eval_num_tokens": 33885773.0,
|
| 1996 |
+
"eval_runtime": 52.0353,
|
| 1997 |
+
"eval_samples_per_second": 22.638,
|
| 1998 |
+
"eval_steps_per_second": 5.669,
|
| 1999 |
+
"step": 7800
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"entropy": 0.048666532533243295,
|
| 2003 |
+
"epoch": 2.2940445743514797,
|
| 2004 |
+
"grad_norm": 0.35318806767463684,
|
| 2005 |
+
"learning_rate": 4.8543884314119306e-05,
|
| 2006 |
+
"loss": 0.05026057243347168,
|
| 2007 |
+
"mean_token_accuracy": 0.9809833306074143,
|
| 2008 |
+
"num_tokens": 34103219.0,
|
| 2009 |
+
"step": 7850
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"entropy": 0.044851412964053455,
|
| 2013 |
+
"epoch": 2.308659115820241,
|
| 2014 |
+
"grad_norm": 0.3726493716239929,
|
| 2015 |
+
"learning_rate": 4.7539666599718824e-05,
|
| 2016 |
+
"loss": 0.042955965995788575,
|
| 2017 |
+
"mean_token_accuracy": 0.9849131184816361,
|
| 2018 |
+
"num_tokens": 34320418.0,
|
| 2019 |
+
"step": 7900
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"entropy": 0.04522866520099342,
|
| 2023 |
+
"epoch": 2.3232736572890027,
|
| 2024 |
+
"grad_norm": 0.27043774724006653,
|
| 2025 |
+
"learning_rate": 4.653544888531834e-05,
|
| 2026 |
+
"loss": 0.046865167617797854,
|
| 2027 |
+
"mean_token_accuracy": 0.9826881098747253,
|
| 2028 |
+
"num_tokens": 34538564.0,
|
| 2029 |
+
"step": 7950
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"entropy": 0.04482917439658195,
|
| 2033 |
+
"epoch": 2.337888198757764,
|
| 2034 |
+
"grad_norm": 0.43633177876472473,
|
| 2035 |
+
"learning_rate": 4.553123117091786e-05,
|
| 2036 |
+
"loss": 0.044361419677734375,
|
| 2037 |
+
"mean_token_accuracy": 0.983955349624157,
|
| 2038 |
+
"num_tokens": 34754870.0,
|
| 2039 |
+
"step": 8000
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 2.337888198757764,
|
| 2043 |
+
"eval_entropy": 0.05105202936141168,
|
| 2044 |
+
"eval_loss": 0.14225418865680695,
|
| 2045 |
+
"eval_mean_token_accuracy": 0.9614736490330453,
|
| 2046 |
+
"eval_num_tokens": 34754870.0,
|
| 2047 |
+
"eval_runtime": 52.0382,
|
| 2048 |
+
"eval_samples_per_second": 22.637,
|
| 2049 |
+
"eval_steps_per_second": 5.669,
|
| 2050 |
+
"step": 8000
|
| 2051 |
+
}
|
| 2052 |
+
],
|
| 2053 |
+
"logging_steps": 50,
|
| 2054 |
+
"max_steps": 10266,
|
| 2055 |
+
"num_input_tokens_seen": 0,
|
| 2056 |
+
"num_train_epochs": 3,
|
| 2057 |
+
"save_steps": 200,
|
| 2058 |
+
"stateful_callbacks": {
|
| 2059 |
+
"TrainerControl": {
|
| 2060 |
+
"args": {
|
| 2061 |
+
"should_epoch_stop": false,
|
| 2062 |
+
"should_evaluate": false,
|
| 2063 |
+
"should_log": false,
|
| 2064 |
+
"should_save": true,
|
| 2065 |
+
"should_training_stop": false
|
| 2066 |
+
},
|
| 2067 |
+
"attributes": {}
|
| 2068 |
+
}
|
| 2069 |
+
},
|
| 2070 |
+
"total_flos": 8.613426021666648e+17,
|
| 2071 |
+
"train_batch_size": 4,
|
| 2072 |
+
"trial_name": null,
|
| 2073 |
+
"trial_params": null
|
| 2074 |
+
}
|
checkpoint-8000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a62e955b4a049523c45a1102ae6e2f9b52bf056eeb0fc0b8d413459417f3ac39
|
| 3 |
+
size 5841
|