Upload MistralForCausalLM
Browse files- config.json +5 -5
- model.safetensors +2 -2
    	
        config.json
    CHANGED
    
    | @@ -4,21 +4,21 @@ | |
| 4 | 
             
              ],
         | 
| 5 | 
             
              "attention_dropout": 0.0,
         | 
| 6 | 
             
              "bos_token_id": 1,
         | 
| 7 | 
            -
              "dtype": " | 
| 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":  | 
| 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- | 
| 20 | 
            -
              "rope_theta":  | 
| 21 | 
            -
              "sliding_window":  | 
| 22 | 
             
              "tie_word_embeddings": false,
         | 
| 23 | 
             
              "transformers_version": "4.57.0.dev0",
         | 
| 24 | 
             
              "use_cache": true,
         | 
|  | |
| 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": 131072,
         | 
| 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-06,
         | 
| 20 | 
            +
              "rope_theta": 10000.0,
         | 
| 21 | 
            +
              "sliding_window": 4096,
         | 
| 22 | 
             
              "tie_word_embeddings": false,
         | 
| 23 | 
             
              "transformers_version": "4.57.0.dev0",
         | 
| 24 | 
             
              "use_cache": true,
         | 
    	
        model.safetensors
    CHANGED
    
    | @@ -1,3 +1,3 @@ | |
| 1 | 
             
            version https://git-lfs.github.com/spec/v1
         | 
| 2 | 
            -
            oid sha256: | 
| 3 | 
            -
            size  | 
|  | |
| 1 | 
             
            version https://git-lfs.github.com/spec/v1
         | 
| 2 | 
            +
            oid sha256:39bf11c8289dda2a5ea0c0ca0dcb1c3831cd5cc81f40668d49e8e26a26beaf0c
         | 
| 3 | 
            +
            size 2058000
         | 

