qgallouedec HF Staff commited on
Commit
40584f5
·
verified ·
1 Parent(s): 872803d

Upload MistralForCausalLM

Browse files
Files changed (3) hide show
  1. config.json +7 -7
  2. generation_config.json +1 -1
  3. model.safetensors +2 -2
config.json CHANGED
@@ -4,23 +4,23 @@
4
  ],
5
  "attention_dropout": 0.0,
6
  "bos_token_id": 1,
 
7
  "eos_token_id": 2,
8
  "head_dim": null,
9
  "hidden_act": "silu",
10
  "hidden_size": 8,
11
  "initializer_range": 0.02,
12
  "intermediate_size": 32,
13
- "max_position_embeddings": 131072,
14
  "model_type": "mistral",
15
  "num_attention_heads": 4,
16
  "num_hidden_layers": 2,
17
  "num_key_value_heads": 2,
18
- "rms_norm_eps": 1e-06,
19
- "rope_theta": 10000.0,
20
- "sliding_window": 4096,
21
  "tie_word_embeddings": false,
22
- "torch_dtype": "float32",
23
- "transformers_version": "4.55.0.dev0",
24
  "use_cache": true,
25
- "vocab_size": 32003
26
  }
 
4
  ],
5
  "attention_dropout": 0.0,
6
  "bos_token_id": 1,
7
+ "dtype": "float32",
8
  "eos_token_id": 2,
9
  "head_dim": null,
10
  "hidden_act": "silu",
11
  "hidden_size": 8,
12
  "initializer_range": 0.02,
13
  "intermediate_size": 32,
14
+ "max_position_embeddings": 32768,
15
  "model_type": "mistral",
16
  "num_attention_heads": 4,
17
  "num_hidden_layers": 2,
18
  "num_key_value_heads": 2,
19
+ "rms_norm_eps": 1e-05,
20
+ "rope_theta": 1000000.0,
21
+ "sliding_window": null,
22
  "tie_word_embeddings": false,
23
+ "transformers_version": "4.57.0.dev0",
 
24
  "use_cache": true,
25
+ "vocab_size": 32000
26
  }
generation_config.json CHANGED
@@ -2,5 +2,5 @@
2
  "_from_model_config": true,
3
  "bos_token_id": 1,
4
  "eos_token_id": 2,
5
- "transformers_version": "4.55.0.dev0"
6
  }
 
2
  "_from_model_config": true,
3
  "bos_token_id": 1,
4
  "eos_token_id": 2,
5
+ "transformers_version": "4.57.0.dev0"
6
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:acf936bf2553fa2e74662332fb763c2c4d27228151419c375598b4a6bd672455
3
- size 2058192
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d60c0b49ae52bab32a5a497cde8d1b63c91062f7a97775e87bb0bf92257858f8
3
+ size 2058000