Hugging Face's logo

impuneetg
/
gpt2-wikitext2

Text Generation
Transformers
TensorBoard
Safetensors
gpt2
Generated from Trainer
text-generation-inference
Model card Files
xet
Metrics Community
impuneetg commited on
Commit
024c5c8
·
verified ·
1 Parent(s): 66376d2

Training in progress, step 500

Browse files
Files changed (13) hide show
  1. config.json +9 -34
  2. model.safetensors +2 -2
  3. runs/Jun02_11-00-27_ip-10-192-11-99/events.out.tfevents.1717326028.ip-10-192-11-99.3471.0 +3 -0
  4. runs/Jun02_11-00-55_ip-10-192-11-99/events.out.tfevents.1717326056.ip-10-192-11-99.3471.1 +3 -0
  5. runs/Jun02_11-01-55_ip-10-192-11-99/events.out.tfevents.1717326119.ip-10-192-11-99.3471.2 +3 -0
  6. runs/Jun02_11-03-07_ip-10-192-11-99/events.out.tfevents.1717326188.ip-10-192-11-99.3471.3 +3 -0
  7. runs/Jun02_11-04-48_ip-10-192-11-99/events.out.tfevents.1717326290.ip-10-192-11-99.3471.4 +3 -0
  8. runs/Jun02_11-38-24_ip-10-192-11-99/events.out.tfevents.1717328307.ip-10-192-11-99.3471.5 +3 -0
  9. runs/Jun02_11-38-44_ip-10-192-11-99/events.out.tfevents.1717328324.ip-10-192-11-99.3471.6 +3 -0
  10. runs/Jun02_11-46-20_ip-10-192-11-99/events.out.tfevents.1717328781.ip-10-192-11-99.3471.7 +3 -0
  11. runs/Jun02_12-17-44_ip-10-192-11-99/events.out.tfevents.1717330665.ip-10-192-11-99.3471.8 +3 -0
  12. runs/Jun02_12-19-09_ip-10-192-11-99/events.out.tfevents.1717330750.ip-10-192-11-99.3471.9 +3 -0
  13. training_args.bin +1 -1
config.json CHANGED
@@ -1,39 +1,14 @@
1
  {
2
- "_name_or_path": "gpt2",
3
- "activation_function": "gelu_new",
4
  "architectures": [
5
- "GPT2LMHeadModel"
6
  ],
7
- "attn_pdrop": 0.1,
8
- "bos_token_id": 0,
9
- "embd_pdrop": 0.1,
10
- "eos_token_id": 0,
11
- "initializer_range": 0.02,
12
- "layer_norm_epsilon": 1e-05,
13
- "model_type": "gpt2",
14
- "n_ctx": 128,
15
- "n_embd": 768,
16
- "n_head": 12,
17
- "n_inner": null,
18
- "n_layer": 3,
19
- "n_positions": 1024,
20
- "reorder_and_upcast_attn": false,
21
- "resid_pdrop": 0.1,
22
- "scale_attn_by_inverse_layer_idx": false,
23
- "scale_attn_weights": true,
24
- "summary_activation": null,
25
- "summary_first_dropout": 0.1,
26
- "summary_proj_to_labels": true,
27
- "summary_type": "cls_index",
28
- "summary_use_proj": true,
29
- "task_specific_params": {
30
- "text-generation": {
31
- "do_sample": true,
32
- "max_length": 50
33
- }
34
- },
35
  "torch_dtype": "float32",
36
- "transformers_version": "4.41.2",
37
- "use_cache": true,
38
- "vocab_size": 25001
39
  }
 
1
  {
 
 
2
  "architectures": [
3
+ "LangModel"
4
  ],
5
+ "d_hid": 200,
6
+ "dropout": 0.2,
7
+ "emsize": 200,
8
+ "model_type": "langmodel",
9
+ "nhead": 2,
10
+ "nlayers": 2,
11
+ "ntokens": 25001,
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
12
  "torch_dtype": "float32",
13
+ "transformers_version": "4.41.2"
 
 
14
  }
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:d74b006d1cb7ab609e212b7fec404394876fdc75dde43cf582cf054f6f19c4eb
3
- size 165013384
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fd6b02ec360cc8a43501031501e6f6082a27d54e5bb684e8966b62ab102d3cf5
3
+ size 46040892
runs/Jun02_11-00-27_ip-10-192-11-99/events.out.tfevents.1717326028.ip-10-192-11-99.3471.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2ea911878fb824da8c0e536f6e8f7b1029f06eaa24861a896b63355c6f860572
3
+ size 4680
runs/Jun02_11-00-55_ip-10-192-11-99/events.out.tfevents.1717326056.ip-10-192-11-99.3471.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59aa4886a4ead5a0ce7b9b1e100fafb59ae19cc85a6eb65bc43938ad40c375a7
3
+ size 7268
runs/Jun02_11-01-55_ip-10-192-11-99/events.out.tfevents.1717326119.ip-10-192-11-99.3471.2 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a9e021936d795bfabc189f378abb75c8dad79be106b680b908a7b6503858e27
3
+ size 4753
runs/Jun02_11-03-07_ip-10-192-11-99/events.out.tfevents.1717326188.ip-10-192-11-99.3471.3 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:859ad1c9e61f878cda97aacb56d4b9a24b6d49df96b638b04218c12f32123929
3
+ size 6853
runs/Jun02_11-04-48_ip-10-192-11-99/events.out.tfevents.1717326290.ip-10-192-11-99.3471.4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73532a187400e008d324fe15b3dcf40604283301e18686d67efcc368d20a4ebb
3
+ size 6851
runs/Jun02_11-38-24_ip-10-192-11-99/events.out.tfevents.1717328307.ip-10-192-11-99.3471.5 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c0ad870f25984140f39b2721429e107d73ce1800ba1ae40a156d0231ea787370
3
+ size 4327
runs/Jun02_11-38-44_ip-10-192-11-99/events.out.tfevents.1717328324.ip-10-192-11-99.3471.6 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa022a1d6ab0ef343849c33dec50970907cea592481a50aca130fb9e9d57b4d0
3
+ size 5115
runs/Jun02_11-46-20_ip-10-192-11-99/events.out.tfevents.1717328781.ip-10-192-11-99.3471.7 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ff1cd6278d4f723f5f5b0c9475e22971c183405a1fa311bdc8338cdf5b816fe
3
+ size 5115
runs/Jun02_12-17-44_ip-10-192-11-99/events.out.tfevents.1717330665.ip-10-192-11-99.3471.8 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:806cd39d1a6ae9d2b7d70bd5297a1d83298fdd70121d60d20e2de6b299265fcd
3
+ size 5115
runs/Jun02_12-19-09_ip-10-192-11-99/events.out.tfevents.1717330750.ip-10-192-11-99.3471.9 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de6f59a150d1308f835f04651a080b110428aee56166bfc15c1def304bb5fd4a
3
+ size 4761
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:2a65e786694ac354bd1a95e717e22d3a472cd9f07b3525111635395fe2ac0ce7
3
  size 5112
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9dc981a91c8b89c8583f01b6a40eac00b48df1dc7c5928f95cc5d0696f93bb2b
3
  size 5112