danielv835 commited on
Commit
e46e29b
1 Parent(s): 1adeed9

Training in progress, step 500

Browse files
.gitignore ADDED
@@ -0,0 +1 @@
 
 
1
+ checkpoint-*/
config.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "bigcode/santacoder",
3
+ "activation_function": "gelu_fast",
4
+ "architectures": [
5
+ "GPT2LMHeadCustomModel"
6
+ ],
7
+ "attention_head_type": "multiquery",
8
+ "attn_pdrop": 0.1,
9
+ "auto_map": {
10
+ "AutoConfig": "configuration_gpt2_mq.GPT2CustomConfig",
11
+ "AutoModelForCausalLM": "modeling_gpt2_mq.GPT2LMHeadCustomModel"
12
+ },
13
+ "bos_token_id": 50256,
14
+ "embd_pdrop": 0.1,
15
+ "eos_token_id": 50256,
16
+ "initializer_range": 0.02,
17
+ "layer_norm_epsilon": 1e-05,
18
+ "model_type": "gpt2",
19
+ "n_embd": 2048,
20
+ "n_head": 16,
21
+ "n_inner": 8192,
22
+ "n_layer": 24,
23
+ "n_positions": 2048,
24
+ "reorder_and_upcast_attn": false,
25
+ "resid_pdrop": 0.1,
26
+ "scale_attn_by_inverse_layer_idx": false,
27
+ "scale_attn_weights": true,
28
+ "summary_activation": null,
29
+ "summary_first_dropout": 0.1,
30
+ "summary_proj_to_labels": true,
31
+ "summary_type": "cls_index",
32
+ "summary_use_proj": true,
33
+ "torch_dtype": "float32",
34
+ "transformers_version": "4.26.1",
35
+ "use_cache": false,
36
+ "vocab_size": 49280
37
+ }
pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:83757ae6f0c8f00f6f84a56a0ddafadb4f562c6e33b6f583526759eed3b943b3
3
+ size 4600336581
runs/Feb15_23-38-40_172cd01d832d/1676504398.765223/events.out.tfevents.1676504398.172cd01d832d.111.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bea506c320ff764794fe2763733e62eb6469e90cb1818d3bfd99102ad6ee2d2c
3
+ size 5722
runs/Feb15_23-38-40_172cd01d832d/events.out.tfevents.1676504398.172cd01d832d.111.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ee68432e7cf09e51fd92051ca06fc4c01332921f40453f9cef994affbd6263b
3
+ size 12299
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf1306e0a1fa3eb98448d1b05dcac1c058faa174d4982c3df82a73c47b8a7f47
3
+ size 3515