appvoid commited on
Commit
cc6e3ef
·
verified ·
1 Parent(s): 9e09cf7

Upload folder using huggingface_hub

Browse files
README.md ADDED
@@ -0,0 +1,52 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model:
3
+ - OrionLLM/Terminus-LFM2.5-350m
4
+ - squ11z1/claude-oss-350m
5
+ - MihaiPopa-1/LFM2.5-350M-heretic
6
+ library_name: transformers
7
+ tags:
8
+ - mergekit
9
+ - merge
10
+
11
+ ---
12
+ # merge
13
+
14
+ This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
15
+
16
+ ## Merge Details
17
+ ### Merge Method
18
+
19
+ This model was merged using the [DARE TIES](https://arxiv.org/abs/2311.03099) merge method using [MihaiPopa-1/LFM2.5-350M-heretic](https://huggingface.co/MihaiPopa-1/LFM2.5-350M-heretic) as a base.
20
+
21
+ ### Models Merged
22
+
23
+ The following models were included in the merge:
24
+ * [OrionLLM/Terminus-LFM2.5-350m](https://huggingface.co/OrionLLM/Terminus-LFM2.5-350m)
25
+ * [squ11z1/claude-oss-350m](https://huggingface.co/squ11z1/claude-oss-350m)
26
+
27
+ ### Configuration
28
+
29
+ The following YAML configuration was used to produce this model:
30
+
31
+ ```yaml
32
+ merge_method: dare_ties
33
+ base_model: MihaiPopa-1/LFM2.5-350M-heretic
34
+ dtype: bfloat16
35
+
36
+ parameters:
37
+ normalize: true
38
+ int8_mask: true
39
+
40
+ models:
41
+ # Main complementary skill: terminal / agentic / tool behavior
42
+ - model: OrionLLM/Terminus-LFM2.5-350m
43
+ parameters:
44
+ weight: 0.31
45
+ density: 0.40
46
+
47
+ # Assistant tone / conversational instruction-following
48
+ - model: squ11z1/claude-oss-350m
49
+ parameters:
50
+ weight: 0.24
51
+ density: 0.36
52
+ ```
config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Lfm2ForCausalLM"
4
+ ],
5
+ "block_auto_adjust_ff_dim": true,
6
+ "block_dim": 1024,
7
+ "block_ffn_dim_multiplier": 1.0,
8
+ "block_mlp_init_scale": 1.0,
9
+ "block_multiple_of": 256,
10
+ "block_norm_eps": 1e-05,
11
+ "block_out_init_scale": 1.0,
12
+ "block_use_swiglu": true,
13
+ "block_use_xavier_init": true,
14
+ "bos_token_id": 1,
15
+ "conv_L_cache": 3,
16
+ "conv_bias": false,
17
+ "conv_dim": 1024,
18
+ "conv_use_xavier_init": true,
19
+ "dtype": "bfloat16",
20
+ "eos_token_id": 7,
21
+ "hidden_size": 1024,
22
+ "initializer_range": 0.02,
23
+ "intermediate_size": 6656,
24
+ "layer_types": [
25
+ "conv",
26
+ "conv",
27
+ "full_attention",
28
+ "conv",
29
+ "conv",
30
+ "full_attention",
31
+ "conv",
32
+ "conv",
33
+ "full_attention",
34
+ "conv",
35
+ "full_attention",
36
+ "conv",
37
+ "full_attention",
38
+ "conv",
39
+ "full_attention",
40
+ "conv"
41
+ ],
42
+ "max_position_embeddings": 128000,
43
+ "model_type": "lfm2",
44
+ "norm_eps": 1e-05,
45
+ "num_attention_heads": 16,
46
+ "num_heads": 16,
47
+ "num_hidden_layers": 16,
48
+ "num_key_value_heads": 8,
49
+ "pad_token_id": 0,
50
+ "rope_parameters": {
51
+ "rope_theta": 1000000.0,
52
+ "rope_type": "default"
53
+ },
54
+ "rope_theta": 1000000.0,
55
+ "transformers_version": "4.57.6",
56
+ "use_cache": true,
57
+ "use_pos_enc": true,
58
+ "vocab_size": 65536
59
+ }
generation_config.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 7,
5
+ "pad_token_id": 0,
6
+ "transformers_version": "5.5.3"
7
+ }
mergekit_config.yml ADDED
@@ -0,0 +1,20 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ merge_method: dare_ties
2
+ base_model: MihaiPopa-1/LFM2.5-350M-heretic
3
+ dtype: bfloat16
4
+
5
+ parameters:
6
+ normalize: true
7
+ int8_mask: true
8
+
9
+ models:
10
+ # Main complementary skill: terminal / agentic / tool behavior
11
+ - model: OrionLLM/Terminus-LFM2.5-350m
12
+ parameters:
13
+ weight: 0.31
14
+ density: 0.40
15
+
16
+ # Assistant tone / conversational instruction-following
17
+ - model: squ11z1/claude-oss-350m
18
+ parameters:
19
+ weight: 0.24
20
+ density: 0.36
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ec3dbbb9c7e9ded5e079233630c71d930510e0efe710aead042b183722355e1c
3
+ size 708984464
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,21 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "backend": "tokenizers",
3
+ "bos_token": "<|startoftext|>",
4
+ "clean_up_tokenization_spaces": false,
5
+ "eos_token": "<|im_end|>",
6
+ "extra_special_tokens": [],
7
+ "is_local": false,
8
+ "legacy": false,
9
+ "model_input_names": [
10
+ "input_ids",
11
+ "attention_mask"
12
+ ],
13
+ "model_max_length": 1000000000000000019884624838656,
14
+ "model_specific_special_tokens": {},
15
+ "pad_token": "<|pad|>",
16
+ "sp_model_kwargs": {},
17
+ "spaces_between_special_tokens": false,
18
+ "tokenizer_class": "TokenizersBackend",
19
+ "use_default_system_prompt": false,
20
+ "use_fast": true
21
+ }