zelc commited on
Commit
20c554d
·
verified ·
1 Parent(s): 444e750

Upload folder using huggingface_hub

Browse files
semionpolicy/epoch_0_step_9000/config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLMEagle3"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": 151643,
8
+ "draft_vocab_size": 32000,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 151645,
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 5120,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 25600,
16
+ "max_position_embeddings": 40960,
17
+ "max_window_layers": 64,
18
+ "mlp_bias": false,
19
+ "model_type": "llama",
20
+ "num_attention_heads": 64,
21
+ "num_hidden_layers": 1,
22
+ "num_key_value_heads": 8,
23
+ "pretraining_tp": 1,
24
+ "rms_norm_eps": 1e-06,
25
+ "rope_scaling": null,
26
+ "rope_theta": 1000000,
27
+ "sliding_window": null,
28
+ "tie_word_embeddings": false,
29
+ "transformers_version": "4.57.1",
30
+ "use_cache": true,
31
+ "use_sliding_window": false,
32
+ "vocab_size": 151936
33
+ }
semionpolicy/epoch_0_step_9000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f781b3f0011e07653b229ceaabe02ae9b1b64fc8fd6db15af8294c986682cba2
3
+ size 1565450112
semionpolicy/epoch_0_step_9000/training_state.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac6cc62cf618755193ddf4b615917b9113ef49eb217e912896853aa998d8e942
3
+ size 4759