pnevskaiaan commited on
Commit
36e44e4
·
verified ·
1 Parent(s): 075a143

Upload 7 files

Browse files
Files changed (3) hide show
  1. config.json +28 -4
  2. generation_config.json +2 -2
  3. pytorch_model.bin +2 -2
config.json CHANGED
@@ -3,31 +3,55 @@
3
  "Lfm2ForCausalLM"
4
  ],
5
  "block_auto_adjust_ff_dim": true,
 
 
6
  "block_ffn_dim_multiplier": 1.0,
 
7
  "block_multiple_of": 8,
 
 
 
 
8
  "bos_token_id": 1,
9
  "conv_L_cache": 3,
10
  "conv_bias": false,
11
- "eos_token_id": 2,
 
 
 
12
  "hidden_size": 16,
13
  "initializer_range": 0.02,
14
  "intermediate_size": 12,
15
  "layer_types": [
 
 
16
  "full_attention",
 
 
17
  "full_attention",
 
 
18
  "full_attention",
19
- "full_attention"
 
 
 
 
 
 
20
  ],
21
  "max_position_embeddings": 128000,
22
  "model_type": "lfm2",
23
  "norm_eps": 1e-05,
24
  "num_attention_heads": 4,
 
25
  "num_hidden_layers": 4,
26
  "num_key_value_heads": 4,
27
  "pad_token_id": 0,
28
  "rope_theta": 1000000.0,
29
- "torch_dtype": "float32",
30
- "transformers_version": "4.55.4",
31
  "use_cache": true,
 
32
  "vocab_size": 65536
33
  }
 
3
  "Lfm2ForCausalLM"
4
  ],
5
  "block_auto_adjust_ff_dim": true,
6
+ "block_dim": 1024,
7
+ "block_ff_dim": 6656,
8
  "block_ffn_dim_multiplier": 1.0,
9
+ "block_mlp_init_scale": 1.0,
10
  "block_multiple_of": 8,
11
+ "block_norm_eps": 1e-05,
12
+ "block_out_init_scale": 1.0,
13
+ "block_use_swiglu": true,
14
+ "block_use_xavier_init": true,
15
  "bos_token_id": 1,
16
  "conv_L_cache": 3,
17
  "conv_bias": false,
18
+ "conv_dim": 1024,
19
+ "conv_dim_out": 1024,
20
+ "conv_use_xavier_init": true,
21
+ "eos_token_id": 7,
22
  "hidden_size": 16,
23
  "initializer_range": 0.02,
24
  "intermediate_size": 12,
25
  "layer_types": [
26
+ "conv",
27
+ "conv",
28
  "full_attention",
29
+ "conv",
30
+ "conv",
31
  "full_attention",
32
+ "conv",
33
+ "conv",
34
  "full_attention",
35
+ "conv",
36
+ "full_attention",
37
+ "conv",
38
+ "full_attention",
39
+ "conv",
40
+ "full_attention",
41
+ "conv"
42
  ],
43
  "max_position_embeddings": 128000,
44
  "model_type": "lfm2",
45
  "norm_eps": 1e-05,
46
  "num_attention_heads": 4,
47
+ "num_heads": 16,
48
  "num_hidden_layers": 4,
49
  "num_key_value_heads": 4,
50
  "pad_token_id": 0,
51
  "rope_theta": 1000000.0,
52
+ "torch_dtype": "bfloat16",
53
+ "transformers_version": "4.54.0",
54
  "use_cache": true,
55
+ "use_pos_enc": true,
56
  "vocab_size": 65536
57
  }
generation_config.json CHANGED
@@ -1,7 +1,7 @@
1
  {
2
  "_from_model_config": true,
3
  "bos_token_id": 1,
4
- "eos_token_id": 2,
5
  "pad_token_id": 0,
6
- "transformers_version": "4.55.4"
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": "4.54.0"
7
  }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:396d9f677e05a8840053d4773a3db96231ad60d674dd7f4d097d987c1e532b03
3
- size 4232657
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d75e7bc19ebd90d1d0ca09566231788bdb098789b5f03ac59725ca28d0eb995e
3
+ size 2120950