Hugging Face's logo

soumitsr
/
smolLM2-article-digestor-4bit

Text Generation
Transformers
Safetensors
English
llama
text-generation-inference
unsloth
trl
sft
conversational
Inference Endpoints
4-bit precision
bitsandbytes
Model card Files Community
soumitsr commited on
Commit
191b7cb
·
verified ·
1 Parent(s): 54fb747

Trained with Unsloth

Browse files

Upload model trained with Unsloth 2x faster

Files changed (3) hide show
  1. config.json +19 -13
  2. generation_config.json +3 -3
  3. model.safetensors +2 -2
config.json CHANGED
@@ -1,28 +1,28 @@
1
  {
2
- "_name_or_path": "unsloth/SmolLM2-135M",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
6
  "attention_bias": false,
7
  "attention_dropout": 0.0,
8
- "bos_token_id": 0,
9
- "eos_token_id": 0,
10
  "head_dim": 64,
11
  "hidden_act": "silu",
12
- "hidden_size": 576,
13
- "initializer_range": 0.041666666666666664,
14
- "intermediate_size": 1536,
15
  "is_llama_config": true,
16
  "max_position_embeddings": 16384,
17
  "mlp_bias": false,
18
  "model_type": "llama",
19
- "num_attention_heads": 9,
20
- "num_hidden_layers": 30,
21
- "num_key_value_heads": 3,
22
  "pad_token_id": 49152,
23
  "pretraining_tp": 1,
24
  "quantization_config": {
25
- "bnb_4bit_compute_dtype": "float16",
26
  "bnb_4bit_quant_type": "nf4",
27
  "bnb_4bit_use_double_quant": true,
28
  "llm_int8_enable_fp32_cpu_offload": false,
@@ -41,9 +41,15 @@
41
  },
42
  "rope_theta": 100000,
43
  "tie_word_embeddings": true,
44
- "torch_dtype": "float16",
45
- "transformers_version": "4.46.2",
46
- "unsloth_version": "2024.11.9",
 
 
 
 
 
 
47
  "use_cache": true,
48
  "vocab_size": 49153
49
  }
 
1
  {
2
+ "_name_or_path": "unsloth/SmolLM2-360M-Instruct",
3
  "architectures": [
4
  "LlamaForCausalLM"
5
  ],
6
  "attention_bias": false,
7
  "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "eos_token_id": 2,
10
  "head_dim": 64,
11
  "hidden_act": "silu",
12
+ "hidden_size": 960,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 2560,
15
  "is_llama_config": true,
16
  "max_position_embeddings": 16384,
17
  "mlp_bias": false,
18
  "model_type": "llama",
19
+ "num_attention_heads": 15,
20
+ "num_hidden_layers": 32,
21
+ "num_key_value_heads": 5,
22
  "pad_token_id": 49152,
23
  "pretraining_tp": 1,
24
  "quantization_config": {
25
+ "bnb_4bit_compute_dtype": "bfloat16",
26
  "bnb_4bit_quant_type": "nf4",
27
  "bnb_4bit_use_double_quant": true,
28
  "llm_int8_enable_fp32_cpu_offload": false,
 
41
  },
42
  "rope_theta": 100000,
43
  "tie_word_embeddings": true,
44
+ "torch_dtype": "bfloat16",
45
+ "transformers.js_config": {
46
+ "kv_cache_dtype": {
47
+ "fp16": "float16",
48
+ "q4f16": "float16"
49
+ }
50
+ },
51
+ "transformers_version": "4.46.3",
52
+ "unsloth_version": "2024.12.4",
53
  "use_cache": true,
54
  "vocab_size": 49153
55
  }
generation_config.json CHANGED
@@ -1,8 +1,8 @@
1
  {
2
  "_from_model_config": true,
3
- "bos_token_id": 0,
4
- "eos_token_id": 0,
5
  "max_length": 16384,
6
  "pad_token_id": 49152,
7
- "transformers_version": "4.46.2"
8
  }
 
1
  {
2
  "_from_model_config": true,
3
+ "bos_token_id": 1,
4
+ "eos_token_id": 2,
5
  "max_length": 16384,
6
  "pad_token_id": 49152,
7
+ "transformers_version": "4.46.3"
8
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:ca04c8d41ac77f07ab7c970e08a1b6b8dd8436e295ca6fee1c95ce8da4e615e1
3
- size 111880641
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ecc8a0a8fbd67b622c91a67658c66f497631945a9268c62f77bbae1cf451e60c
3
+ size 257223517