NowaBwagel0 commited on
Commit
ad3b272
1 Parent(s): 9a463a1

Training in progress, step 500

Browse files
config.json CHANGED
@@ -1,29 +1,30 @@
1
- {
2
- "_name_or_path": "NowaBwagel0/llama-68m-oasst",
3
- "architectures": [
4
- "LlamaForCausalLM"
5
- ],
6
- "attention_bias": false,
7
- "attention_dropout": 0.0,
8
- "bos_token_id": 0,
9
- "eos_token_id": 2,
10
- "hidden_act": "silu",
11
- "hidden_size": 768,
12
- "initializer_range": 0.02,
13
- "intermediate_size": 3072,
14
- "max_position_embeddings": 2048,
15
- "model_type": "llama",
16
- "num_attention_heads": 12,
17
- "num_hidden_layers": 2,
18
- "num_key_value_heads": 12,
19
- "pad_token_id": 1,
20
- "pretraining_tp": 1,
21
- "rms_norm_eps": 1e-06,
22
- "rope_scaling": null,
23
- "rope_theta": 10000.0,
24
- "tie_word_embeddings": false,
25
- "torch_dtype": "float32",
26
- "transformers_version": "4.36.2",
27
- "use_cache": true,
28
- "vocab_size": 32000
29
- }
 
 
1
+ {
2
+ "_name_or_path": "NowaBwagel0/llama-68m-oasst",
3
+ "architectures": [
4
+ "LlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 0,
9
+ "eos_token_id": 2,
10
+ "hidden_act": "silu",
11
+ "hidden_size": 768,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 3072,
14
+ "max_position_embeddings": 2048,
15
+ "mlp_bias": false,
16
+ "model_type": "llama",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 2,
19
+ "num_key_value_heads": 12,
20
+ "pad_token_id": 1,
21
+ "pretraining_tp": 1,
22
+ "rms_norm_eps": 1e-06,
23
+ "rope_scaling": null,
24
+ "rope_theta": 10000.0,
25
+ "tie_word_embeddings": false,
26
+ "torch_dtype": "float32",
27
+ "transformers_version": "4.41.2",
28
+ "use_cache": true,
29
+ "vocab_size": 32000
30
+ }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:222fd55c2572308839069ef9cac45f9f1e021306ca818d359402018a09225ddf
3
  size 272123144
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6402be19a463d5ec383596ef98c4b6464888a4d6c9eed95c3ef34e8e29c54e67
3
  size 272123144
runs/Jul06_15-19-14_Noah-Desktop/events.out.tfevents.1720297154.Noah-Desktop.9832.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c23c8f9124e5156dd505acf8efb0493313e5c5d05c0761733f5c442a14d0206e
3
+ size 4158
runs/Jul06_15-21-23_Noah-Desktop/events.out.tfevents.1720297284.Noah-Desktop.19916.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:33198d15368e48df38977858901373d453c356d17a1c3273ac7beda53fc4dae5
3
+ size 18192
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:eef0aecfcb87afb8a469d4f3011f493d11544b599c76bd1f1057ec6438d0aada
3
- size 4728
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:074fe31fee23eaf64ee8376c4f76c01955d2e1b6945c939dc79070a2717497dc
3
+ size 5112