Alex-xu commited on
Commit
f5ea9cc
1 Parent(s): e144934

Training in progress, step 2000, checkpoint

Browse files
checkpoint-2000/config.json CHANGED
@@ -16,7 +16,7 @@
16
  "intermediate_size": 3072,
17
  "layer_norm_eps": 1e-12,
18
  "max_blocks": 200,
19
- "max_position_embeddings": 1801,
20
  "max_relative_position_embeddings": 8,
21
  "model_type": "longelm",
22
  "node_size": 1,
 
16
  "intermediate_size": 3072,
17
  "layer_norm_eps": 1e-12,
18
  "max_blocks": 200,
19
+ "max_position_embeddings": 1805,
20
  "max_relative_position_embeddings": 8,
21
  "model_type": "longelm",
22
  "node_size": 1,
checkpoint-2000/optimizer.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6eb49ee6d613e8282121ae998cde2ee47490fd0c77710aae3f46f02be6337b7a
3
- size 1010407418
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f413f554ea3cdedfe4ea911dadb0a97bed7a78d2771d7305e9054ae620985852
3
+ size 1010431994
checkpoint-2000/pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a9ef586c497792cc4fa858bb7c3fbbcbec4a2928ef9570d5457b4342f04ee9f0
3
- size 505188394
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d5d740497db60e67b7729e13cb35e8385072aa1ff787a4904f8a4a4413db5fa9
3
+ size 505200682
checkpoint-2000/trainer_state.json CHANGED
@@ -9,35 +9,35 @@
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
- "ep_loss": 4.7343,
13
  "epoch": 0.04,
14
  "learning_rate": 2.5e-05,
15
- "loss": 11.6218,
16
- "mlm_loss": 6.8875,
17
  "step": 500
18
  },
19
  {
20
- "ep_loss": 0.4315,
21
  "epoch": 0.08,
22
  "learning_rate": 5e-05,
23
- "loss": 3.0285,
24
- "mlm_loss": 2.5969,
25
  "step": 1000
26
  },
27
  {
28
- "ep_loss": 0.3811,
29
  "epoch": 0.12,
30
  "learning_rate": 7.5e-05,
31
- "loss": 2.2571,
32
- "mlm_loss": 1.8761,
33
  "step": 1500
34
  },
35
  {
36
- "ep_loss": 0.37,
37
  "epoch": 0.15,
38
  "learning_rate": 0.0001,
39
- "loss": 1.9695,
40
- "mlm_loss": 1.5994,
41
  "step": 2000
42
  }
43
  ],
 
9
  "is_world_process_zero": true,
10
  "log_history": [
11
  {
12
+ "ep_loss": 4.4238,
13
  "epoch": 0.04,
14
  "learning_rate": 2.5e-05,
15
+ "loss": 11.1467,
16
+ "mlm_loss": 6.7229,
17
  "step": 500
18
  },
19
  {
20
+ "ep_loss": 0.4305,
21
  "epoch": 0.08,
22
  "learning_rate": 5e-05,
23
+ "loss": 2.9841,
24
+ "mlm_loss": 2.5537,
25
  "step": 1000
26
  },
27
  {
28
+ "ep_loss": 0.3724,
29
  "epoch": 0.12,
30
  "learning_rate": 7.5e-05,
31
+ "loss": 2.2371,
32
+ "mlm_loss": 1.8647,
33
  "step": 1500
34
  },
35
  {
36
+ "ep_loss": 0.364,
37
  "epoch": 0.15,
38
  "learning_rate": 0.0001,
39
+ "loss": 1.8477,
40
+ "mlm_loss": 1.4837,
41
  "step": 2000
42
  }
43
  ],
checkpoint-2000/training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cee56d4f38339eac5ecf098381d5222b3783455863547a17ac0df67c8caa7d15
3
  size 4664
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b26e03be70ebe5d40a81a81be94421c0578a9d8742b065bc029501df48632957
3
  size 4664