Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- checkpoints-v5.5/checkpoint-16384/ema.safetensors +3 -0
- checkpoints-v5.5/checkpoint-16384/eval_state.json +3 -0
- checkpoints-v5.5/checkpoint-16384/model.safetensors +3 -0
- checkpoints-v5.5/checkpoint-16384/optimizer.pt +3 -0
- checkpoints-v5.5/checkpoint-16384/rng_state.pth +3 -0
- checkpoints-v5.5/checkpoint-16384/scaler.pt +3 -0
- checkpoints-v5.5/checkpoint-16384/scheduler.pt +3 -0
- checkpoints-v5.5/checkpoint-16384/trainer_state.json +642 -0
- checkpoints-v5.5/checkpoint-16384/training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -39,3 +39,4 @@ checkpoints-v5.2/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs
|
|
| 39 |
checkpoints-v5.3/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 40 |
checkpoints-v5.4/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 41 |
checkpoints-v5.5/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 39 |
checkpoints-v5.3/checkpoint-11264/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 40 |
checkpoints-v5.4/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 41 |
checkpoints-v5.5/checkpoint-10240/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 42 |
+
checkpoints-v5.5/checkpoint-16384/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
checkpoints-v5.5/checkpoint-16384/ema.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b214552f793aa63ee0d330804edee200bb7c6bfe8bbc6636adae6937a1d7085b
|
| 3 |
+
size 54599592
|
checkpoints-v5.5/checkpoint-16384/eval_state.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f338225d77eeb5095f6d4f56632459411a4f5d635de3925c7f8b9e176e6446fe
|
| 3 |
+
size 19624464
|
checkpoints-v5.5/checkpoint-16384/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c4c58e94e19d3b8b3b229e31035a91edfeee2d4df93751fab3b27a7e2ae24ad7
|
| 3 |
+
size 54599624
|
checkpoints-v5.5/checkpoint-16384/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f223d0fddad74363f94a0cd0e79a73d5dc7ef42068e5574353ff9e7ec036b6d5
|
| 3 |
+
size 76551435
|
checkpoints-v5.5/checkpoint-16384/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e4fa627e98670eb5920d7b317183b0266ad722cdbb1470b10882f9343b56c790
|
| 3 |
+
size 14645
|
checkpoints-v5.5/checkpoint-16384/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6066d02a5b2336fdede61162351e70da932528611ee67fafecfcfe4f1fbbb67c
|
| 3 |
+
size 1383
|
checkpoints-v5.5/checkpoint-16384/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c9df748f8c7ad998decb1236c5f0350ad44e146e99d94157849c4c744f6fd5c1
|
| 3 |
+
size 1465
|
checkpoints-v5.5/checkpoint-16384/trainer_state.json
ADDED
|
@@ -0,0 +1,642 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.11817182752602,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 16384,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.13239381989786023,
|
| 14 |
+
"grad_norm": 7.3368072509765625,
|
| 15 |
+
"learning_rate": 9.990234375e-05,
|
| 16 |
+
"loss": 13.833250999450684,
|
| 17 |
+
"step": 1024
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.13239381989786023,
|
| 21 |
+
"eval_bleu": 0.08843010093383347,
|
| 22 |
+
"eval_ce_clean_loss": 3.785046512154257,
|
| 23 |
+
"eval_ce_pred_loss": 6.1898151750017885,
|
| 24 |
+
"eval_flow_cos_loss": 0.39729490059955863,
|
| 25 |
+
"eval_flow_mse_loss": 1.2504585077808161,
|
| 26 |
+
"eval_loss": 9.467699330323821,
|
| 27 |
+
"flow/cos_sim": 0.6027051043358578,
|
| 28 |
+
"flow/improvement_ratio": 0.9944390654563904,
|
| 29 |
+
"flow/mag_ratio_mean": 0.6029561613775363,
|
| 30 |
+
"flow/mag_ratio_std": 0.06967356720357944,
|
| 31 |
+
"step": 1024
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.13239381989786023,
|
| 35 |
+
"eval_bleu": 0.08843010093383347,
|
| 36 |
+
"eval_ce_clean_loss": 3.785046512154257,
|
| 37 |
+
"eval_ce_pred_loss": 6.1898151750017885,
|
| 38 |
+
"eval_flow_cos_loss": 0.39729490059955863,
|
| 39 |
+
"eval_flow_mse_loss": 1.2504585077808161,
|
| 40 |
+
"eval_loss": 9.467699330323821,
|
| 41 |
+
"eval_runtime": 69.1426,
|
| 42 |
+
"eval_samples_per_second": 144.629,
|
| 43 |
+
"eval_steps_per_second": 2.271,
|
| 44 |
+
"flow/cos_sim": 0.6027051043358578,
|
| 45 |
+
"flow/improvement_ratio": 0.9944390654563904,
|
| 46 |
+
"flow/mag_ratio_mean": 0.6029561613775363,
|
| 47 |
+
"flow/mag_ratio_std": 0.06967356720357944,
|
| 48 |
+
"step": 1024
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.26478763979572045,
|
| 52 |
+
"grad_norm": 1.5976208448410034,
|
| 53 |
+
"learning_rate": 9.971175203561169e-05,
|
| 54 |
+
"loss": 6.556396484375,
|
| 55 |
+
"step": 2048
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.26478763979572045,
|
| 59 |
+
"eval_bleu": 0.2647666813502405,
|
| 60 |
+
"eval_ce_clean_loss": 0.6921462336922907,
|
| 61 |
+
"eval_ce_pred_loss": 4.159270631279915,
|
| 62 |
+
"eval_flow_cos_loss": 0.33016856043202103,
|
| 63 |
+
"eval_flow_mse_loss": 1.083006814786583,
|
| 64 |
+
"eval_loss": 4.769184565088551,
|
| 65 |
+
"flow/cos_sim": 0.6698314437441005,
|
| 66 |
+
"flow/improvement_ratio": 0.9946717445258122,
|
| 67 |
+
"flow/mag_ratio_mean": 0.6456947994839614,
|
| 68 |
+
"flow/mag_ratio_std": 0.08488734332239552,
|
| 69 |
+
"step": 2048
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.26478763979572045,
|
| 73 |
+
"eval_bleu": 0.2647666813502405,
|
| 74 |
+
"eval_ce_clean_loss": 0.6921462336922907,
|
| 75 |
+
"eval_ce_pred_loss": 4.159270631279915,
|
| 76 |
+
"eval_flow_cos_loss": 0.33016856043202103,
|
| 77 |
+
"eval_flow_mse_loss": 1.083006814786583,
|
| 78 |
+
"eval_loss": 4.769184565088551,
|
| 79 |
+
"eval_runtime": 67.8586,
|
| 80 |
+
"eval_samples_per_second": 147.365,
|
| 81 |
+
"eval_steps_per_second": 2.314,
|
| 82 |
+
"flow/cos_sim": 0.6698314437441005,
|
| 83 |
+
"flow/improvement_ratio": 0.9946717445258122,
|
| 84 |
+
"flow/mag_ratio_mean": 0.6456947994839614,
|
| 85 |
+
"flow/mag_ratio_std": 0.08488734332239552,
|
| 86 |
+
"step": 2048
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.3971814596935807,
|
| 90 |
+
"grad_norm": 1.0349175930023193,
|
| 91 |
+
"learning_rate": 9.885033161800567e-05,
|
| 92 |
+
"loss": 4.22868013381958,
|
| 93 |
+
"step": 3072
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.3971814596935807,
|
| 97 |
+
"eval_bleu": 0.3307477170153146,
|
| 98 |
+
"eval_ce_clean_loss": 0.21776276523140586,
|
| 99 |
+
"eval_ce_pred_loss": 3.5308286536271405,
|
| 100 |
+
"eval_flow_cos_loss": 0.3201853291244264,
|
| 101 |
+
"eval_flow_mse_loss": 1.0728299655732076,
|
| 102 |
+
"eval_loss": 3.8422190869689747,
|
| 103 |
+
"flow/cos_sim": 0.6798147354156349,
|
| 104 |
+
"flow/improvement_ratio": 0.9928556031482235,
|
| 105 |
+
"flow/mag_ratio_mean": 0.6495526474752243,
|
| 106 |
+
"flow/mag_ratio_std": 0.08990857878308388,
|
| 107 |
+
"step": 3072
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.3971814596935807,
|
| 111 |
+
"eval_bleu": 0.3307477170153146,
|
| 112 |
+
"eval_ce_clean_loss": 0.21776276523140586,
|
| 113 |
+
"eval_ce_pred_loss": 3.5308286536271405,
|
| 114 |
+
"eval_flow_cos_loss": 0.3201853291244264,
|
| 115 |
+
"eval_flow_mse_loss": 1.0728299655732076,
|
| 116 |
+
"eval_loss": 3.8422190869689747,
|
| 117 |
+
"eval_runtime": 67.9151,
|
| 118 |
+
"eval_samples_per_second": 147.243,
|
| 119 |
+
"eval_steps_per_second": 2.312,
|
| 120 |
+
"flow/cos_sim": 0.6798147354156349,
|
| 121 |
+
"flow/improvement_ratio": 0.9928556031482235,
|
| 122 |
+
"flow/mag_ratio_mean": 0.6495526474752243,
|
| 123 |
+
"flow/mag_ratio_std": 0.08990857878308388,
|
| 124 |
+
"step": 3072
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.5295752795914409,
|
| 128 |
+
"grad_norm": 2.284677028656006,
|
| 129 |
+
"learning_rate": 9.742400750550229e-05,
|
| 130 |
+
"loss": 3.7094979286193848,
|
| 131 |
+
"step": 4096
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.5295752795914409,
|
| 135 |
+
"eval_bleu": 0.3527590985954599,
|
| 136 |
+
"eval_ce_clean_loss": 0.10325595935818496,
|
| 137 |
+
"eval_ce_pred_loss": 3.3037764221240002,
|
| 138 |
+
"eval_flow_cos_loss": 0.29567099870390195,
|
| 139 |
+
"eval_flow_mse_loss": 1.0334580939286833,
|
| 140 |
+
"eval_loss": 3.5232752614719853,
|
| 141 |
+
"flow/cos_sim": 0.7043290168616423,
|
| 142 |
+
"flow/improvement_ratio": 0.9942314150227103,
|
| 143 |
+
"flow/mag_ratio_mean": 0.6710645951283206,
|
| 144 |
+
"flow/mag_ratio_std": 0.08415729860970929,
|
| 145 |
+
"step": 4096
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.5295752795914409,
|
| 149 |
+
"eval_bleu": 0.3527590985954599,
|
| 150 |
+
"eval_ce_clean_loss": 0.10325595935818496,
|
| 151 |
+
"eval_ce_pred_loss": 3.3037764221240002,
|
| 152 |
+
"eval_flow_cos_loss": 0.29567099870390195,
|
| 153 |
+
"eval_flow_mse_loss": 1.0334580939286833,
|
| 154 |
+
"eval_loss": 3.5232752614719853,
|
| 155 |
+
"eval_runtime": 67.527,
|
| 156 |
+
"eval_samples_per_second": 148.089,
|
| 157 |
+
"eval_steps_per_second": 2.325,
|
| 158 |
+
"flow/cos_sim": 0.7043290168616423,
|
| 159 |
+
"flow/improvement_ratio": 0.9942314150227103,
|
| 160 |
+
"flow/mag_ratio_mean": 0.6710645951283206,
|
| 161 |
+
"flow/mag_ratio_std": 0.08415729860970929,
|
| 162 |
+
"step": 4096
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 0.6619690994893012,
|
| 166 |
+
"grad_norm": 1.0625219345092773,
|
| 167 |
+
"learning_rate": 9.544981995345226e-05,
|
| 168 |
+
"loss": 3.4675893783569336,
|
| 169 |
+
"step": 5120
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.6619690994893012,
|
| 173 |
+
"eval_bleu": 0.37785369114831263,
|
| 174 |
+
"eval_ce_clean_loss": 0.0570975638404014,
|
| 175 |
+
"eval_ce_pred_loss": 3.1086114409622874,
|
| 176 |
+
"eval_flow_cos_loss": 0.2890773550340324,
|
| 177 |
+
"eval_flow_mse_loss": 1.0530509029983715,
|
| 178 |
+
"eval_loss": 3.3584457916818606,
|
| 179 |
+
"flow/cos_sim": 0.7109226474336757,
|
| 180 |
+
"flow/improvement_ratio": 0.9944721566643685,
|
| 181 |
+
"flow/mag_ratio_mean": 0.6796372863137798,
|
| 182 |
+
"flow/mag_ratio_std": 0.08220032687969268,
|
| 183 |
+
"step": 5120
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.6619690994893012,
|
| 187 |
+
"eval_bleu": 0.37785369114831263,
|
| 188 |
+
"eval_ce_clean_loss": 0.0570975638404014,
|
| 189 |
+
"eval_ce_pred_loss": 3.1086114409622874,
|
| 190 |
+
"eval_flow_cos_loss": 0.2890773550340324,
|
| 191 |
+
"eval_flow_mse_loss": 1.0530509029983715,
|
| 192 |
+
"eval_loss": 3.3584457916818606,
|
| 193 |
+
"eval_runtime": 67.5831,
|
| 194 |
+
"eval_samples_per_second": 147.966,
|
| 195 |
+
"eval_steps_per_second": 2.323,
|
| 196 |
+
"flow/cos_sim": 0.7109226474336757,
|
| 197 |
+
"flow/improvement_ratio": 0.9944721566643685,
|
| 198 |
+
"flow/mag_ratio_mean": 0.6796372863137798,
|
| 199 |
+
"flow/mag_ratio_std": 0.08220032687969268,
|
| 200 |
+
"step": 5120
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.7943629193871614,
|
| 204 |
+
"grad_norm": 1.390210747718811,
|
| 205 |
+
"learning_rate": 9.295057566334431e-05,
|
| 206 |
+
"loss": 3.3179638385772705,
|
| 207 |
+
"step": 6144
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.7943629193871614,
|
| 211 |
+
"eval_bleu": 0.40284924593911503,
|
| 212 |
+
"eval_ce_clean_loss": 0.03560329580999863,
|
| 213 |
+
"eval_ce_pred_loss": 2.935079864635589,
|
| 214 |
+
"eval_flow_cos_loss": 0.2695348363393431,
|
| 215 |
+
"eval_flow_mse_loss": 1.037700882383213,
|
| 216 |
+
"eval_loss": 3.195243762556914,
|
| 217 |
+
"flow/cos_sim": 0.7304651706841341,
|
| 218 |
+
"flow/improvement_ratio": 0.9945835288922498,
|
| 219 |
+
"flow/mag_ratio_mean": 0.7005888793119199,
|
| 220 |
+
"flow/mag_ratio_std": 0.0835665136480787,
|
| 221 |
+
"step": 6144
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.7943629193871614,
|
| 225 |
+
"eval_bleu": 0.40284924593911503,
|
| 226 |
+
"eval_ce_clean_loss": 0.03560329580999863,
|
| 227 |
+
"eval_ce_pred_loss": 2.935079864635589,
|
| 228 |
+
"eval_flow_cos_loss": 0.2695348363393431,
|
| 229 |
+
"eval_flow_mse_loss": 1.037700882383213,
|
| 230 |
+
"eval_loss": 3.195243762556914,
|
| 231 |
+
"eval_runtime": 67.6818,
|
| 232 |
+
"eval_samples_per_second": 147.75,
|
| 233 |
+
"eval_steps_per_second": 2.32,
|
| 234 |
+
"flow/cos_sim": 0.7304651706841341,
|
| 235 |
+
"flow/improvement_ratio": 0.9945835288922498,
|
| 236 |
+
"flow/mag_ratio_mean": 0.7005888793119199,
|
| 237 |
+
"flow/mag_ratio_std": 0.0835665136480787,
|
| 238 |
+
"step": 6144
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.9267567392850217,
|
| 242 |
+
"grad_norm": 1.7730882167816162,
|
| 243 |
+
"learning_rate": 8.995830349195804e-05,
|
| 244 |
+
"loss": 3.2147014141082764,
|
| 245 |
+
"step": 7168
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 0.9267567392850217,
|
| 249 |
+
"eval_bleu": 0.4108774093856199,
|
| 250 |
+
"eval_ce_clean_loss": 0.025004512588879105,
|
| 251 |
+
"eval_ce_pred_loss": 2.8478709542827243,
|
| 252 |
+
"eval_flow_cos_loss": 0.25457756676871307,
|
| 253 |
+
"eval_flow_mse_loss": 1.0318333833080948,
|
| 254 |
+
"eval_loss": 3.113991931745201,
|
| 255 |
+
"flow/cos_sim": 0.745422419469068,
|
| 256 |
+
"flow/improvement_ratio": 0.9949804590006542,
|
| 257 |
+
"flow/mag_ratio_mean": 0.7171509364607987,
|
| 258 |
+
"flow/mag_ratio_std": 0.08899391105600224,
|
| 259 |
+
"step": 7168
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 0.9267567392850217,
|
| 263 |
+
"eval_bleu": 0.4108774093856199,
|
| 264 |
+
"eval_ce_clean_loss": 0.025004512588879105,
|
| 265 |
+
"eval_ce_pred_loss": 2.8478709542827243,
|
| 266 |
+
"eval_flow_cos_loss": 0.25457756676871307,
|
| 267 |
+
"eval_flow_mse_loss": 1.0318333833080948,
|
| 268 |
+
"eval_loss": 3.113991931745201,
|
| 269 |
+
"eval_runtime": 67.7148,
|
| 270 |
+
"eval_samples_per_second": 147.678,
|
| 271 |
+
"eval_steps_per_second": 2.319,
|
| 272 |
+
"flow/cos_sim": 0.745422419469068,
|
| 273 |
+
"flow/improvement_ratio": 0.9949804590006542,
|
| 274 |
+
"flow/mag_ratio_mean": 0.7171509364607987,
|
| 275 |
+
"flow/mag_ratio_std": 0.08899391105600224,
|
| 276 |
+
"step": 7168
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.05908591376301,
|
| 280 |
+
"grad_norm": 2.208953619003296,
|
| 281 |
+
"learning_rate": 8.650172716103233e-05,
|
| 282 |
+
"loss": 3.119405508041382,
|
| 283 |
+
"step": 8192
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 1.05908591376301,
|
| 287 |
+
"eval_bleu": 0.4100243274079566,
|
| 288 |
+
"eval_ce_clean_loss": 0.018418982221632248,
|
| 289 |
+
"eval_ce_pred_loss": 2.801309669853016,
|
| 290 |
+
"eval_flow_cos_loss": 0.24036190160520518,
|
| 291 |
+
"eval_flow_mse_loss": 1.0228236701078475,
|
| 292 |
+
"eval_loss": 3.062249883724626,
|
| 293 |
+
"flow/cos_sim": 0.7596381280072935,
|
| 294 |
+
"flow/improvement_ratio": 0.9959618415042852,
|
| 295 |
+
"flow/mag_ratio_mean": 0.7364322915198697,
|
| 296 |
+
"flow/mag_ratio_std": 0.0917528191949152,
|
| 297 |
+
"step": 8192
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.05908591376301,
|
| 301 |
+
"eval_bleu": 0.4100243274079566,
|
| 302 |
+
"eval_ce_clean_loss": 0.018418982221632248,
|
| 303 |
+
"eval_ce_pred_loss": 2.801309669853016,
|
| 304 |
+
"eval_flow_cos_loss": 0.24036190160520518,
|
| 305 |
+
"eval_flow_mse_loss": 1.0228236701078475,
|
| 306 |
+
"eval_loss": 3.062249883724626,
|
| 307 |
+
"eval_runtime": 69.3159,
|
| 308 |
+
"eval_samples_per_second": 144.267,
|
| 309 |
+
"eval_steps_per_second": 2.265,
|
| 310 |
+
"flow/cos_sim": 0.7596381280072935,
|
| 311 |
+
"flow/improvement_ratio": 0.9959618415042852,
|
| 312 |
+
"flow/mag_ratio_mean": 0.7364322915198697,
|
| 313 |
+
"flow/mag_ratio_std": 0.0917528191949152,
|
| 314 |
+
"step": 8192
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 1.19147973366087,
|
| 318 |
+
"grad_norm": 1.0985896587371826,
|
| 319 |
+
"learning_rate": 8.263142386444264e-05,
|
| 320 |
+
"loss": 3.0675039291381836,
|
| 321 |
+
"step": 9216
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 1.19147973366087,
|
| 325 |
+
"eval_bleu": 0.42913809326340735,
|
| 326 |
+
"eval_ce_clean_loss": 0.014243369497904543,
|
| 327 |
+
"eval_ce_pred_loss": 2.679083149903899,
|
| 328 |
+
"eval_flow_cos_loss": 0.2306169035138598,
|
| 329 |
+
"eval_flow_mse_loss": 1.0239125793906534,
|
| 330 |
+
"eval_loss": 2.97116835861449,
|
| 331 |
+
"flow/cos_sim": 0.7693831127160674,
|
| 332 |
+
"flow/improvement_ratio": 0.9939722635184124,
|
| 333 |
+
"flow/mag_ratio_mean": 0.748531128190885,
|
| 334 |
+
"flow/mag_ratio_std": 0.0947970964821281,
|
| 335 |
+
"step": 9216
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 1.19147973366087,
|
| 339 |
+
"eval_bleu": 0.42913809326340735,
|
| 340 |
+
"eval_ce_clean_loss": 0.014243369497904543,
|
| 341 |
+
"eval_ce_pred_loss": 2.679083149903899,
|
| 342 |
+
"eval_flow_cos_loss": 0.2306169035138598,
|
| 343 |
+
"eval_flow_mse_loss": 1.0239125793906534,
|
| 344 |
+
"eval_loss": 2.97116835861449,
|
| 345 |
+
"eval_runtime": 67.1108,
|
| 346 |
+
"eval_samples_per_second": 149.007,
|
| 347 |
+
"eval_steps_per_second": 2.339,
|
| 348 |
+
"flow/cos_sim": 0.7693831127160674,
|
| 349 |
+
"flow/improvement_ratio": 0.9939722635184124,
|
| 350 |
+
"flow/mag_ratio_mean": 0.748531128190885,
|
| 351 |
+
"flow/mag_ratio_std": 0.0947970964821281,
|
| 352 |
+
"step": 9216
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.3238735535587303,
|
| 356 |
+
"grad_norm": 2.3166847229003906,
|
| 357 |
+
"learning_rate": 7.837697175482903e-05,
|
| 358 |
+
"loss": 3.002436399459839,
|
| 359 |
+
"step": 10240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 1.3238735535587303,
|
| 363 |
+
"eval_bleu": 0.42768151912552244,
|
| 364 |
+
"eval_ce_clean_loss": 0.011440879840308883,
|
| 365 |
+
"eval_ce_pred_loss": 2.672361337455215,
|
| 366 |
+
"eval_flow_cos_loss": 0.2179829450739417,
|
| 367 |
+
"eval_flow_mse_loss": 1.0098259748926588,
|
| 368 |
+
"eval_loss": 2.946415506350766,
|
| 369 |
+
"flow/cos_sim": 0.782017103426016,
|
| 370 |
+
"flow/improvement_ratio": 0.9939965146362402,
|
| 371 |
+
"flow/mag_ratio_mean": 0.7625711093283003,
|
| 372 |
+
"flow/mag_ratio_std": 0.09684707447411908,
|
| 373 |
+
"step": 10240
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 1.3238735535587303,
|
| 377 |
+
"eval_bleu": 0.42768151912552244,
|
| 378 |
+
"eval_ce_clean_loss": 0.011440879840308883,
|
| 379 |
+
"eval_ce_pred_loss": 2.672361337455215,
|
| 380 |
+
"eval_flow_cos_loss": 0.2179829450739417,
|
| 381 |
+
"eval_flow_mse_loss": 1.0098259748926588,
|
| 382 |
+
"eval_loss": 2.946415506350766,
|
| 383 |
+
"eval_runtime": 67.7743,
|
| 384 |
+
"eval_samples_per_second": 147.549,
|
| 385 |
+
"eval_steps_per_second": 2.317,
|
| 386 |
+
"flow/cos_sim": 0.782017103426016,
|
| 387 |
+
"flow/improvement_ratio": 0.9939965146362402,
|
| 388 |
+
"flow/mag_ratio_mean": 0.7625711093283003,
|
| 389 |
+
"flow/mag_ratio_std": 0.09684707447411908,
|
| 390 |
+
"step": 10240
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 1.4562673734565905,
|
| 394 |
+
"grad_norm": 1.1153963804244995,
|
| 395 |
+
"learning_rate": 7.37946961193838e-05,
|
| 396 |
+
"loss": 2.948077917098999,
|
| 397 |
+
"step": 11264
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 1.4562673734565905,
|
| 401 |
+
"eval_bleu": 0.44608815072120206,
|
| 402 |
+
"eval_ce_clean_loss": 0.009407547670327554,
|
| 403 |
+
"eval_ce_pred_loss": 2.5671096637750126,
|
| 404 |
+
"eval_flow_cos_loss": 0.21284512444666237,
|
| 405 |
+
"eval_flow_mse_loss": 1.0222040597040942,
|
| 406 |
+
"eval_loss": 2.8817996310580307,
|
| 407 |
+
"flow/cos_sim": 0.7871549156061404,
|
| 408 |
+
"flow/improvement_ratio": 0.9953155885836121,
|
| 409 |
+
"flow/mag_ratio_mean": 0.767739696867147,
|
| 410 |
+
"flow/mag_ratio_std": 0.10089007670142848,
|
| 411 |
+
"step": 11264
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.4562673734565905,
|
| 415 |
+
"eval_bleu": 0.44608815072120206,
|
| 416 |
+
"eval_ce_clean_loss": 0.009407547670327554,
|
| 417 |
+
"eval_ce_pred_loss": 2.5671096637750126,
|
| 418 |
+
"eval_flow_cos_loss": 0.21284512444666237,
|
| 419 |
+
"eval_flow_mse_loss": 1.0222040597040942,
|
| 420 |
+
"eval_loss": 2.8817996310580307,
|
| 421 |
+
"eval_runtime": 69.2843,
|
| 422 |
+
"eval_samples_per_second": 144.333,
|
| 423 |
+
"eval_steps_per_second": 2.266,
|
| 424 |
+
"flow/cos_sim": 0.7871549156061404,
|
| 425 |
+
"flow/improvement_ratio": 0.9953155885836121,
|
| 426 |
+
"flow/mag_ratio_mean": 0.767739696867147,
|
| 427 |
+
"flow/mag_ratio_std": 0.10089007670142848,
|
| 428 |
+
"step": 11264
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.5886611933544508,
|
| 432 |
+
"grad_norm": 2.7540619373321533,
|
| 433 |
+
"learning_rate": 6.894239286709331e-05,
|
| 434 |
+
"loss": 2.921593427658081,
|
| 435 |
+
"step": 12288
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 1.5886611933544508,
|
| 439 |
+
"eval_bleu": 0.4600711751490207,
|
| 440 |
+
"eval_ce_clean_loss": 0.008017947172129135,
|
| 441 |
+
"eval_ce_pred_loss": 2.4964520749013137,
|
| 442 |
+
"eval_flow_cos_loss": 0.20653787796284742,
|
| 443 |
+
"eval_flow_mse_loss": 1.0247267002512694,
|
| 444 |
+
"eval_loss": 2.8318955564195183,
|
| 445 |
+
"flow/cos_sim": 0.7934621500361497,
|
| 446 |
+
"flow/improvement_ratio": 0.9952654098249545,
|
| 447 |
+
"flow/mag_ratio_mean": 0.7712421458997544,
|
| 448 |
+
"flow/mag_ratio_std": 0.10261066300663979,
|
| 449 |
+
"step": 12288
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 1.5886611933544508,
|
| 453 |
+
"eval_bleu": 0.4600711751490207,
|
| 454 |
+
"eval_ce_clean_loss": 0.008017947172129135,
|
| 455 |
+
"eval_ce_pred_loss": 2.4964520749013137,
|
| 456 |
+
"eval_flow_cos_loss": 0.20653787796284742,
|
| 457 |
+
"eval_flow_mse_loss": 1.0247267002512694,
|
| 458 |
+
"eval_loss": 2.8318955564195183,
|
| 459 |
+
"eval_runtime": 68.3644,
|
| 460 |
+
"eval_samples_per_second": 146.275,
|
| 461 |
+
"eval_steps_per_second": 2.297,
|
| 462 |
+
"flow/cos_sim": 0.7934621500361497,
|
| 463 |
+
"flow/improvement_ratio": 0.9952654098249545,
|
| 464 |
+
"flow/mag_ratio_mean": 0.7712421458997544,
|
| 465 |
+
"flow/mag_ratio_std": 0.10261066300663979,
|
| 466 |
+
"step": 12288
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.721055013252311,
|
| 470 |
+
"grad_norm": 1.0973964929580688,
|
| 471 |
+
"learning_rate": 6.386664071821704e-05,
|
| 472 |
+
"loss": 2.8836159706115723,
|
| 473 |
+
"step": 13312
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.721055013252311,
|
| 477 |
+
"eval_bleu": 0.45198542436217626,
|
| 478 |
+
"eval_ce_clean_loss": 0.00690584495562799,
|
| 479 |
+
"eval_ce_pred_loss": 2.526137225946803,
|
| 480 |
+
"eval_flow_cos_loss": 0.1992794095900408,
|
| 481 |
+
"eval_flow_mse_loss": 1.0171066932617479,
|
| 482 |
+
"eval_loss": 2.842128411979432,
|
| 483 |
+
"flow/cos_sim": 0.8007206206868409,
|
| 484 |
+
"flow/improvement_ratio": 0.9942179546234714,
|
| 485 |
+
"flow/mag_ratio_mean": 0.7813193399435395,
|
| 486 |
+
"flow/mag_ratio_std": 0.10161699169570473,
|
| 487 |
+
"step": 13312
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.721055013252311,
|
| 491 |
+
"eval_bleu": 0.45198542436217626,
|
| 492 |
+
"eval_ce_clean_loss": 0.00690584495562799,
|
| 493 |
+
"eval_ce_pred_loss": 2.526137225946803,
|
| 494 |
+
"eval_flow_cos_loss": 0.1992794095900408,
|
| 495 |
+
"eval_flow_mse_loss": 1.0171066932617479,
|
| 496 |
+
"eval_loss": 2.842128411979432,
|
| 497 |
+
"eval_runtime": 67.7624,
|
| 498 |
+
"eval_samples_per_second": 147.574,
|
| 499 |
+
"eval_steps_per_second": 2.317,
|
| 500 |
+
"flow/cos_sim": 0.8007206206868409,
|
| 501 |
+
"flow/improvement_ratio": 0.9942179546234714,
|
| 502 |
+
"flow/mag_ratio_mean": 0.7813193399435395,
|
| 503 |
+
"flow/mag_ratio_std": 0.10161699169570473,
|
| 504 |
+
"step": 13312
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 1.8534488331501713,
|
| 508 |
+
"grad_norm": 1.9011768102645874,
|
| 509 |
+
"learning_rate": 5.863069490629029e-05,
|
| 510 |
+
"loss": 2.8486995697021484,
|
| 511 |
+
"step": 14336
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 1.8534488331501713,
|
| 515 |
+
"eval_bleu": 0.4661235270091544,
|
| 516 |
+
"eval_ce_clean_loss": 0.0061916545386408355,
|
| 517 |
+
"eval_ce_pred_loss": 2.4525129415427043,
|
| 518 |
+
"eval_flow_cos_loss": 0.19385759503978073,
|
| 519 |
+
"eval_flow_mse_loss": 1.014179768835663,
|
| 520 |
+
"eval_loss": 2.785594868811832,
|
| 521 |
+
"flow/cos_sim": 0.8061424322948334,
|
| 522 |
+
"flow/improvement_ratio": 0.9953108915857448,
|
| 523 |
+
"flow/mag_ratio_mean": 0.7884186832768143,
|
| 524 |
+
"flow/mag_ratio_std": 0.10213647872968844,
|
| 525 |
+
"step": 14336
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 1.8534488331501713,
|
| 529 |
+
"eval_bleu": 0.4661235270091544,
|
| 530 |
+
"eval_ce_clean_loss": 0.0061916545386408355,
|
| 531 |
+
"eval_ce_pred_loss": 2.4525129415427043,
|
| 532 |
+
"eval_flow_cos_loss": 0.19385759503978073,
|
| 533 |
+
"eval_flow_mse_loss": 1.014179768835663,
|
| 534 |
+
"eval_loss": 2.785594868811832,
|
| 535 |
+
"eval_runtime": 66.8958,
|
| 536 |
+
"eval_samples_per_second": 149.486,
|
| 537 |
+
"eval_steps_per_second": 2.347,
|
| 538 |
+
"flow/cos_sim": 0.8061424322948334,
|
| 539 |
+
"flow/improvement_ratio": 0.9953108915857448,
|
| 540 |
+
"flow/mag_ratio_mean": 0.7884186832768143,
|
| 541 |
+
"flow/mag_ratio_std": 0.10213647872968844,
|
| 542 |
+
"step": 14336
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.9858426530480315,
|
| 546 |
+
"grad_norm": 1.449341893196106,
|
| 547 |
+
"learning_rate": 5.330028268696606e-05,
|
| 548 |
+
"loss": 2.8369064331054688,
|
| 549 |
+
"step": 15360
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.9858426530480315,
|
| 553 |
+
"eval_bleu": 0.47279436776678946,
|
| 554 |
+
"eval_ce_clean_loss": 0.005674782390024062,
|
| 555 |
+
"eval_ce_pred_loss": 2.4236644658313433,
|
| 556 |
+
"eval_flow_cos_loss": 0.18836158210304893,
|
| 557 |
+
"eval_flow_mse_loss": 1.0086946066018123,
|
| 558 |
+
"eval_loss": 2.7580248884334684,
|
| 559 |
+
"flow/cos_sim": 0.8116384444722704,
|
| 560 |
+
"flow/improvement_ratio": 0.9943355352255949,
|
| 561 |
+
"flow/mag_ratio_mean": 0.7959212903763838,
|
| 562 |
+
"flow/mag_ratio_std": 0.10561748859798832,
|
| 563 |
+
"step": 15360
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.9858426530480315,
|
| 567 |
+
"eval_bleu": 0.47279436776678946,
|
| 568 |
+
"eval_ce_clean_loss": 0.005674782390024062,
|
| 569 |
+
"eval_ce_pred_loss": 2.4236644658313433,
|
| 570 |
+
"eval_flow_cos_loss": 0.18836158210304893,
|
| 571 |
+
"eval_flow_mse_loss": 1.0086946066018123,
|
| 572 |
+
"eval_loss": 2.7580248884334684,
|
| 573 |
+
"eval_runtime": 67.2307,
|
| 574 |
+
"eval_samples_per_second": 148.742,
|
| 575 |
+
"eval_steps_per_second": 2.335,
|
| 576 |
+
"flow/cos_sim": 0.8116384444722704,
|
| 577 |
+
"flow/improvement_ratio": 0.9943355352255949,
|
| 578 |
+
"flow/mag_ratio_mean": 0.7959212903763838,
|
| 579 |
+
"flow/mag_ratio_std": 0.10561748859798832,
|
| 580 |
+
"step": 15360
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.11817182752602,
|
| 584 |
+
"grad_norm": 3.0230202674865723,
|
| 585 |
+
"learning_rate": 4.792657229246779e-05,
|
| 586 |
+
"loss": 2.816126585006714,
|
| 587 |
+
"step": 16384
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.11817182752602,
|
| 591 |
+
"eval_bleu": 0.47827420818893907,
|
| 592 |
+
"eval_ce_clean_loss": 0.005247315107162591,
|
| 593 |
+
"eval_ce_pred_loss": 2.396967076951531,
|
| 594 |
+
"eval_flow_cos_loss": 0.18433132226679735,
|
| 595 |
+
"eval_flow_mse_loss": 1.0063348123981695,
|
| 596 |
+
"eval_loss": 2.735541887344069,
|
| 597 |
+
"flow/cos_sim": 0.8156687302194583,
|
| 598 |
+
"flow/improvement_ratio": 0.9953372854336052,
|
| 599 |
+
"flow/mag_ratio_mean": 0.7989175444955279,
|
| 600 |
+
"flow/mag_ratio_std": 0.10495429419598003,
|
| 601 |
+
"step": 16384
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.11817182752602,
|
| 605 |
+
"eval_bleu": 0.47827420818893907,
|
| 606 |
+
"eval_ce_clean_loss": 0.005247315107162591,
|
| 607 |
+
"eval_ce_pred_loss": 2.396967076951531,
|
| 608 |
+
"eval_flow_cos_loss": 0.18433132226679735,
|
| 609 |
+
"eval_flow_mse_loss": 1.0063348123981695,
|
| 610 |
+
"eval_loss": 2.735541887344069,
|
| 611 |
+
"eval_runtime": 67.2185,
|
| 612 |
+
"eval_samples_per_second": 148.769,
|
| 613 |
+
"eval_steps_per_second": 2.336,
|
| 614 |
+
"flow/cos_sim": 0.8156687302194583,
|
| 615 |
+
"flow/improvement_ratio": 0.9953372854336052,
|
| 616 |
+
"flow/mag_ratio_mean": 0.7989175444955279,
|
| 617 |
+
"flow/mag_ratio_std": 0.10495429419598003,
|
| 618 |
+
"step": 16384
|
| 619 |
+
}
|
| 620 |
+
],
|
| 621 |
+
"logging_steps": 1024,
|
| 622 |
+
"max_steps": 30940,
|
| 623 |
+
"num_input_tokens_seen": 0,
|
| 624 |
+
"num_train_epochs": 4,
|
| 625 |
+
"save_steps": 1024,
|
| 626 |
+
"stateful_callbacks": {
|
| 627 |
+
"TrainerControl": {
|
| 628 |
+
"args": {
|
| 629 |
+
"should_epoch_stop": false,
|
| 630 |
+
"should_evaluate": false,
|
| 631 |
+
"should_log": false,
|
| 632 |
+
"should_save": true,
|
| 633 |
+
"should_training_stop": false
|
| 634 |
+
},
|
| 635 |
+
"attributes": {}
|
| 636 |
+
}
|
| 637 |
+
},
|
| 638 |
+
"total_flos": 0.0,
|
| 639 |
+
"train_batch_size": 64,
|
| 640 |
+
"trial_name": null,
|
| 641 |
+
"trial_params": null
|
| 642 |
+
}
|
checkpoints-v5.5/checkpoint-16384/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7be4d80b8499f3b5f618b042dcec062719328222caddac0d4e4ce11d371480d
|
| 3 |
+
size 5137
|