Hugging Face's logo

Siddharth63
/
LLM2vec-Smol360-mntp

Safetensors
llama
Model card Files
xet
Community
3v324v23 commited on
Commit
7681e47
·
1 Parent(s): c3879e3

initial commit

Browse files
Files changed (3) hide show
  1. README.md +0 -3
  2. config.json +4 -3
  3. model.safetensors +2 -2
README.md DELETED
@@ -1,3 +0,0 @@
1
- ---
2
- license: artistic-2.0
3
- ---
 
 
 
 
config.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "_name_or_path": "HuggingFaceTB/SmolLM2-360M",
3
  "architectures": [
4
- "LlamaModel"
5
  ],
6
  "attention_bias": false,
7
  "attention_dropout": 0.0,
@@ -13,6 +13,7 @@
13
  "intermediate_size": 2560,
14
  "is_llama_config": true,
15
  "max_position_embeddings": 8192,
 
16
  "model_type": "llama",
17
  "num_attention_heads": 15,
18
  "num_hidden_layers": 32,
@@ -23,8 +24,8 @@
23
  "rope_scaling": null,
24
  "rope_theta": 100000,
25
  "tie_word_embeddings": true,
26
- "torch_dtype": "bfloat16",
27
- "transformers_version": "4.40.2",
28
  "use_cache": true,
29
  "vocab_size": 49152
30
  }
 
1
  {
2
  "_name_or_path": "HuggingFaceTB/SmolLM2-360M",
3
  "architectures": [
4
+ "LlamaBiModel"
5
  ],
6
  "attention_bias": false,
7
  "attention_dropout": 0.0,
 
13
  "intermediate_size": 2560,
14
  "is_llama_config": true,
15
  "max_position_embeddings": 8192,
16
+ "mlp_bias": false,
17
  "model_type": "llama",
18
  "num_attention_heads": 15,
19
  "num_hidden_layers": 32,
 
24
  "rope_scaling": null,
25
  "rope_theta": 100000,
26
  "tie_word_embeddings": true,
27
+ "torch_dtype": "float32",
28
+ "transformers_version": "4.44.2",
29
  "use_cache": true,
30
  "vocab_size": 49152
31
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7f25f5987edb5120ee46c009be41b5d0a28c1583e8b591c9e52f7d8e3a6386c5
3
- size 723673176
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:488bd6902ec21953eb09dc0cc291a38f15e1e30ff8486a6430e7beb7736d555b
3
+ size 1447315344