Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- checkpoints-v2.6/checkpoint-12288/eval_state.json +3 -0
- checkpoints-v2.6/checkpoint-12288/model.safetensors +3 -0
- checkpoints-v2.6/checkpoint-12288/optimizer.pt +3 -0
- checkpoints-v2.6/checkpoint-12288/rng_state.pth +3 -0
- checkpoints-v2.6/checkpoint-12288/scaler.pt +3 -0
- checkpoints-v2.6/checkpoint-12288/scheduler.pt +3 -0
- checkpoints-v2.6/checkpoint-12288/trainer_state.json +958 -0
- checkpoints-v2.6/checkpoint-12288/training_args.bin +3 -0
.gitattributes
CHANGED
|
@@ -39,3 +39,4 @@ checkpoints-v2.1/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs
|
|
| 39 |
checkpoints-v2.2/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 40 |
checkpoints-v2.3/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 41 |
checkpoints-v2.5/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 39 |
checkpoints-v2.2/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 40 |
checkpoints-v2.3/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 41 |
checkpoints-v2.5/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
| 42 |
+
checkpoints-v2.6/checkpoint-12288/eval_state.json filter=lfs diff=lfs merge=lfs -text
|
checkpoints-v2.6/checkpoint-12288/eval_state.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:64adffb790cb59a22cb9881c4844dc3bbd3e7155ebfe1e01f8f19faa296ec6f6
|
| 3 |
+
size 44107863
|
checkpoints-v2.6/checkpoint-12288/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1cc0c28b61de6dac74c60e45ad7135a0bbe5e446655a19c7fd4b36261cac23b4
|
| 3 |
+
size 37402680
|
checkpoints-v2.6/checkpoint-12288/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ab51f2a176e646a1010b5f063fb1a1fb5588eeefd32cd561694cb23aad84ba8
|
| 3 |
+
size 512267
|
checkpoints-v2.6/checkpoint-12288/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9f36e6c28b77555b6da6de84681647b558ac8ebc553a1b458e45112e416a213c
|
| 3 |
+
size 14645
|
checkpoints-v2.6/checkpoint-12288/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6fa220a1da697fc6e732f23a692e1fe49da32dad48c61771249ee01a41c4206
|
| 3 |
+
size 1383
|
checkpoints-v2.6/checkpoint-12288/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:87b2b9d7efb9af569cd6999393703c76b185862a0221b931604de9e5bf2b79a6
|
| 3 |
+
size 1465
|
checkpoints-v2.6/checkpoint-12288/trainer_state.json
ADDED
|
@@ -0,0 +1,958 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.5675488430095608,
|
| 6 |
+
"eval_steps": 1024,
|
| 7 |
+
"global_step": 12288,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.011823934229365849,
|
| 14 |
+
"grad_norm": 1.1330132484436035,
|
| 15 |
+
"learning_rate": 1.9615384615384617e-05,
|
| 16 |
+
"loss": 10.3987,
|
| 17 |
+
"step": 256
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.023647868458731697,
|
| 21 |
+
"grad_norm": 1.1089212894439697,
|
| 22 |
+
"learning_rate": 3.930769230769231e-05,
|
| 23 |
+
"loss": 7.8889,
|
| 24 |
+
"step": 512
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.03547180268809755,
|
| 28 |
+
"grad_norm": 0.8628233075141907,
|
| 29 |
+
"learning_rate": 4.999617095521894e-05,
|
| 30 |
+
"loss": 5.5834,
|
| 31 |
+
"step": 768
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.047295736917463395,
|
| 35 |
+
"grad_norm": 0.5602379441261292,
|
| 36 |
+
"learning_rate": 4.9961092368776736e-05,
|
| 37 |
+
"loss": 3.7587,
|
| 38 |
+
"step": 1024
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.047295736917463395,
|
| 42 |
+
"eval_acr_loss": 0.472869769419165,
|
| 43 |
+
"eval_across_var": 0.004283991528007299,
|
| 44 |
+
"eval_bleu": 0.5621407857406262,
|
| 45 |
+
"eval_ce_loss": 2.3296360490528962,
|
| 46 |
+
"eval_cos_loss": 0.9237595144744333,
|
| 47 |
+
"eval_cov": 0.07785117789490582,
|
| 48 |
+
"eval_cov_loss": 0.009674696304907674,
|
| 49 |
+
"eval_glb_loss": 0.3056357570979149,
|
| 50 |
+
"eval_global_kurtosis": 3.016345693640513,
|
| 51 |
+
"eval_global_mean": -0.0019390732730360336,
|
| 52 |
+
"eval_global_var": 0.3470509150256849,
|
| 53 |
+
"eval_krt_loss": 0.00041134808841301906,
|
| 54 |
+
"eval_loss": 2.7735937942652944,
|
| 55 |
+
"eval_mean_loss": 5.7365890913968915e-06,
|
| 56 |
+
"eval_mse_loss": 1.926733713443965,
|
| 57 |
+
"eval_per_loss": 0.3164680907715401,
|
| 58 |
+
"eval_per_var": 0.337350505672089,
|
| 59 |
+
"eval_within_var": 0.342682637784579,
|
| 60 |
+
"eval_wth_loss": 0.310607642547725,
|
| 61 |
+
"step": 1024
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 0.047295736917463395,
|
| 65 |
+
"eval_acr_loss": 0.472869769419165,
|
| 66 |
+
"eval_across_var": 0.004283991528007299,
|
| 67 |
+
"eval_bleu": 0.5621407857406262,
|
| 68 |
+
"eval_ce_loss": 2.3296360490528962,
|
| 69 |
+
"eval_cos_loss": 0.9237595144744333,
|
| 70 |
+
"eval_cov": 0.07785117789490582,
|
| 71 |
+
"eval_cov_loss": 0.009674696304907674,
|
| 72 |
+
"eval_glb_loss": 0.3056357570979149,
|
| 73 |
+
"eval_global_kurtosis": 3.016345693640513,
|
| 74 |
+
"eval_global_mean": -0.0019390732730360336,
|
| 75 |
+
"eval_global_var": 0.3470509150256849,
|
| 76 |
+
"eval_krt_loss": 0.00041134808841301906,
|
| 77 |
+
"eval_loss": 2.7735937942652944,
|
| 78 |
+
"eval_mean_loss": 5.7365890913968915e-06,
|
| 79 |
+
"eval_mse_loss": 1.926733713443965,
|
| 80 |
+
"eval_per_loss": 0.3164680907715401,
|
| 81 |
+
"eval_per_var": 0.337350505672089,
|
| 82 |
+
"eval_runtime": 158.0332,
|
| 83 |
+
"eval_samples_per_second": 177.134,
|
| 84 |
+
"eval_steps_per_second": 2.772,
|
| 85 |
+
"eval_within_var": 0.342682637784579,
|
| 86 |
+
"eval_wth_loss": 0.310607642547725,
|
| 87 |
+
"step": 1024
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.05911967114682925,
|
| 91 |
+
"grad_norm": 0.3792508840560913,
|
| 92 |
+
"learning_rate": 4.988941132556799e-05,
|
| 93 |
+
"loss": 2.6954,
|
| 94 |
+
"step": 1280
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.0709436053761951,
|
| 98 |
+
"grad_norm": 0.30784618854522705,
|
| 99 |
+
"learning_rate": 4.9781232937269974e-05,
|
| 100 |
+
"loss": 2.0719,
|
| 101 |
+
"step": 1536
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.08276753960556095,
|
| 105 |
+
"grad_norm": 0.24569711089134216,
|
| 106 |
+
"learning_rate": 4.963671583455164e-05,
|
| 107 |
+
"loss": 1.664,
|
| 108 |
+
"step": 1792
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.09459147383492679,
|
| 112 |
+
"grad_norm": 0.20954662561416626,
|
| 113 |
+
"learning_rate": 4.945607193446079e-05,
|
| 114 |
+
"loss": 1.385,
|
| 115 |
+
"step": 2048
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.09459147383492679,
|
| 119 |
+
"eval_acr_loss": 0.467126801201742,
|
| 120 |
+
"eval_across_var": 0.014505732470706582,
|
| 121 |
+
"eval_bleu": 0.8102335908319817,
|
| 122 |
+
"eval_ce_loss": 0.6907969389331939,
|
| 123 |
+
"eval_cos_loss": 0.7647272306217995,
|
| 124 |
+
"eval_cov": 0.10773917420269692,
|
| 125 |
+
"eval_cov_loss": 0.017931701802567805,
|
| 126 |
+
"eval_glb_loss": 0.000625664478229841,
|
| 127 |
+
"eval_global_kurtosis": 3.000208995113634,
|
| 128 |
+
"eval_global_mean": -0.007508264829034675,
|
| 129 |
+
"eval_global_var": 0.8768238085045662,
|
| 130 |
+
"eval_krt_loss": 7.712448770311225e-05,
|
| 131 |
+
"eval_loss": 1.0269616710268743,
|
| 132 |
+
"eval_mean_loss": 6.29432441826079e-05,
|
| 133 |
+
"eval_mse_loss": 1.6563664652985526,
|
| 134 |
+
"eval_per_loss": 0.0019870867933301875,
|
| 135 |
+
"eval_per_var": 0.8565188891267124,
|
| 136 |
+
"eval_within_var": 0.8617800931407981,
|
| 137 |
+
"eval_wth_loss": 0.0015375203581266992,
|
| 138 |
+
"step": 2048
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.09459147383492679,
|
| 142 |
+
"eval_acr_loss": 0.467126801201742,
|
| 143 |
+
"eval_across_var": 0.014505732470706582,
|
| 144 |
+
"eval_bleu": 0.8102335908319817,
|
| 145 |
+
"eval_ce_loss": 0.6907969389331939,
|
| 146 |
+
"eval_cos_loss": 0.7647272306217995,
|
| 147 |
+
"eval_cov": 0.10773917420269692,
|
| 148 |
+
"eval_cov_loss": 0.017931701802567805,
|
| 149 |
+
"eval_glb_loss": 0.000625664478229841,
|
| 150 |
+
"eval_global_kurtosis": 3.000208995113634,
|
| 151 |
+
"eval_global_mean": -0.007508264829034675,
|
| 152 |
+
"eval_global_var": 0.8768238085045662,
|
| 153 |
+
"eval_krt_loss": 7.712448770311225e-05,
|
| 154 |
+
"eval_loss": 1.0269616710268743,
|
| 155 |
+
"eval_mean_loss": 6.29432441826079e-05,
|
| 156 |
+
"eval_mse_loss": 1.6563664652985526,
|
| 157 |
+
"eval_per_loss": 0.0019870867933301875,
|
| 158 |
+
"eval_per_var": 0.8565188891267124,
|
| 159 |
+
"eval_runtime": 153.9643,
|
| 160 |
+
"eval_samples_per_second": 181.815,
|
| 161 |
+
"eval_steps_per_second": 2.845,
|
| 162 |
+
"eval_within_var": 0.8617800931407981,
|
| 163 |
+
"eval_wth_loss": 0.0015375203581266992,
|
| 164 |
+
"step": 2048
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.10641540806429264,
|
| 168 |
+
"grad_norm": 0.181256964802742,
|
| 169 |
+
"learning_rate": 4.923956612967301e-05,
|
| 170 |
+
"loss": 1.1836,
|
| 171 |
+
"step": 2304
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.1182393422936585,
|
| 175 |
+
"grad_norm": 0.1728077083826065,
|
| 176 |
+
"learning_rate": 4.898751590005826e-05,
|
| 177 |
+
"loss": 1.024,
|
| 178 |
+
"step": 2560
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.13006327652302435,
|
| 182 |
+
"grad_norm": 0.14891040325164795,
|
| 183 |
+
"learning_rate": 4.870029084713462e-05,
|
| 184 |
+
"loss": 0.9004,
|
| 185 |
+
"step": 2816
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.1418872107523902,
|
| 189 |
+
"grad_norm": 0.12918385863304138,
|
| 190 |
+
"learning_rate": 4.837831215209188e-05,
|
| 191 |
+
"loss": 0.8021,
|
| 192 |
+
"step": 3072
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.1418872107523902,
|
| 196 |
+
"eval_acr_loss": 0.46483140122400574,
|
| 197 |
+
"eval_across_var": 0.01694616893323385,
|
| 198 |
+
"eval_bleu": 0.8960603905933928,
|
| 199 |
+
"eval_ce_loss": 0.32335185586181403,
|
| 200 |
+
"eval_cos_loss": 0.6106928648197487,
|
| 201 |
+
"eval_cov": 0.09281384890482305,
|
| 202 |
+
"eval_cov_loss": 0.013490723856388841,
|
| 203 |
+
"eval_glb_loss": 0.0,
|
| 204 |
+
"eval_global_kurtosis": 3.000047793127086,
|
| 205 |
+
"eval_global_mean": -0.0062043639622866836,
|
| 206 |
+
"eval_global_var": 0.9301077786101598,
|
| 207 |
+
"eval_krt_loss": 0.0001482390632322396,
|
| 208 |
+
"eval_loss": 0.5985009211655621,
|
| 209 |
+
"eval_mean_loss": 4.6196501266608506e-05,
|
| 210 |
+
"eval_mse_loss": 1.3684135288952692,
|
| 211 |
+
"eval_per_loss": 6.070195510882551e-06,
|
| 212 |
+
"eval_per_var": 0.906668049015411,
|
| 213 |
+
"eval_within_var": 0.9128873076068756,
|
| 214 |
+
"eval_wth_loss": 7.710680918438013e-07,
|
| 215 |
+
"step": 3072
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 0.1418872107523902,
|
| 219 |
+
"eval_acr_loss": 0.46483140122400574,
|
| 220 |
+
"eval_across_var": 0.01694616893323385,
|
| 221 |
+
"eval_bleu": 0.8960603905933928,
|
| 222 |
+
"eval_ce_loss": 0.32335185586181403,
|
| 223 |
+
"eval_cos_loss": 0.6106928648197487,
|
| 224 |
+
"eval_cov": 0.09281384890482305,
|
| 225 |
+
"eval_cov_loss": 0.013490723856388841,
|
| 226 |
+
"eval_glb_loss": 0.0,
|
| 227 |
+
"eval_global_kurtosis": 3.000047793127086,
|
| 228 |
+
"eval_global_mean": -0.0062043639622866836,
|
| 229 |
+
"eval_global_var": 0.9301077786101598,
|
| 230 |
+
"eval_krt_loss": 0.0001482390632322396,
|
| 231 |
+
"eval_loss": 0.5985009211655621,
|
| 232 |
+
"eval_mean_loss": 4.6196501266608506e-05,
|
| 233 |
+
"eval_mse_loss": 1.3684135288952692,
|
| 234 |
+
"eval_per_loss": 6.070195510882551e-06,
|
| 235 |
+
"eval_per_var": 0.906668049015411,
|
| 236 |
+
"eval_runtime": 154.7557,
|
| 237 |
+
"eval_samples_per_second": 180.885,
|
| 238 |
+
"eval_steps_per_second": 2.83,
|
| 239 |
+
"eval_within_var": 0.9128873076068756,
|
| 240 |
+
"eval_wth_loss": 7.710680918438013e-07,
|
| 241 |
+
"step": 3072
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.15371114498175603,
|
| 245 |
+
"grad_norm": 0.12574972212314606,
|
| 246 |
+
"learning_rate": 4.802205195817963e-05,
|
| 247 |
+
"loss": 0.7236,
|
| 248 |
+
"step": 3328
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.1655350792111219,
|
| 252 |
+
"grad_norm": 0.12267891317605972,
|
| 253 |
+
"learning_rate": 4.763203267836576e-05,
|
| 254 |
+
"loss": 0.656,
|
| 255 |
+
"step": 3584
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.17735901344048774,
|
| 259 |
+
"grad_norm": 0.10911328345537186,
|
| 260 |
+
"learning_rate": 4.720882622928019e-05,
|
| 261 |
+
"loss": 0.6002,
|
| 262 |
+
"step": 3840
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.18918294766985358,
|
| 266 |
+
"grad_norm": 0.10669343918561935,
|
| 267 |
+
"learning_rate": 4.675305319256765e-05,
|
| 268 |
+
"loss": 0.5541,
|
| 269 |
+
"step": 4096
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.18918294766985358,
|
| 273 |
+
"eval_acr_loss": 0.461427516861049,
|
| 274 |
+
"eval_across_var": 0.02085038360216778,
|
| 275 |
+
"eval_bleu": 0.9355169660485144,
|
| 276 |
+
"eval_ce_loss": 0.18811083091720598,
|
| 277 |
+
"eval_cos_loss": 0.4924817247886092,
|
| 278 |
+
"eval_cov": 0.08738018715218322,
|
| 279 |
+
"eval_cov_loss": 0.011900046778647186,
|
| 280 |
+
"eval_glb_loss": 0.0,
|
| 281 |
+
"eval_global_kurtosis": 3.0061902341232996,
|
| 282 |
+
"eval_global_mean": -0.007297369579201964,
|
| 283 |
+
"eval_global_var": 1.0064546767979452,
|
| 284 |
+
"eval_krt_loss": 0.00036125986233864483,
|
| 285 |
+
"eval_loss": 0.41625422633946213,
|
| 286 |
+
"eval_mean_loss": 6.224468990837522e-05,
|
| 287 |
+
"eval_mse_loss": 1.1413913132393196,
|
| 288 |
+
"eval_per_loss": 0.0,
|
| 289 |
+
"eval_per_var": 0.9814564604737442,
|
| 290 |
+
"eval_within_var": 0.9855326701244808,
|
| 291 |
+
"eval_wth_loss": 0.0,
|
| 292 |
+
"step": 4096
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.18918294766985358,
|
| 296 |
+
"eval_acr_loss": 0.461427516861049,
|
| 297 |
+
"eval_across_var": 0.02085038360216778,
|
| 298 |
+
"eval_bleu": 0.9355169660485144,
|
| 299 |
+
"eval_ce_loss": 0.18811083091720598,
|
| 300 |
+
"eval_cos_loss": 0.4924817247886092,
|
| 301 |
+
"eval_cov": 0.08738018715218322,
|
| 302 |
+
"eval_cov_loss": 0.011900046778647186,
|
| 303 |
+
"eval_glb_loss": 0.0,
|
| 304 |
+
"eval_global_kurtosis": 3.0061902341232996,
|
| 305 |
+
"eval_global_mean": -0.007297369579201964,
|
| 306 |
+
"eval_global_var": 1.0064546767979452,
|
| 307 |
+
"eval_krt_loss": 0.00036125986233864483,
|
| 308 |
+
"eval_loss": 0.41625422633946213,
|
| 309 |
+
"eval_mean_loss": 6.224468990837522e-05,
|
| 310 |
+
"eval_mse_loss": 1.1413913132393196,
|
| 311 |
+
"eval_per_loss": 0.0,
|
| 312 |
+
"eval_per_var": 0.9814564604737442,
|
| 313 |
+
"eval_runtime": 155.5272,
|
| 314 |
+
"eval_samples_per_second": 179.988,
|
| 315 |
+
"eval_steps_per_second": 2.816,
|
| 316 |
+
"eval_within_var": 0.9855326701244808,
|
| 317 |
+
"eval_wth_loss": 0.0,
|
| 318 |
+
"step": 4096
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.20100688189921945,
|
| 322 |
+
"grad_norm": 0.10487012565135956,
|
| 323 |
+
"learning_rate": 4.6265381904878854e-05,
|
| 324 |
+
"loss": 0.5121,
|
| 325 |
+
"step": 4352
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.2128308161285853,
|
| 329 |
+
"grad_norm": 0.09705886989831924,
|
| 330 |
+
"learning_rate": 4.57465274778347e-05,
|
| 331 |
+
"loss": 0.4805,
|
| 332 |
+
"step": 4608
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.22465475035795113,
|
| 336 |
+
"grad_norm": 0.10024245083332062,
|
| 337 |
+
"learning_rate": 4.519725074940068e-05,
|
| 338 |
+
"loss": 0.4469,
|
| 339 |
+
"step": 4864
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.236478684587317,
|
| 343 |
+
"grad_norm": 0.08834318816661835,
|
| 344 |
+
"learning_rate": 4.461835716820895e-05,
|
| 345 |
+
"loss": 0.42,
|
| 346 |
+
"step": 5120
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.236478684587317,
|
| 350 |
+
"eval_acr_loss": 0.4548332993978779,
|
| 351 |
+
"eval_across_var": 0.028460351478087304,
|
| 352 |
+
"eval_bleu": 0.9552258623563333,
|
| 353 |
+
"eval_ce_loss": 0.12340653128071463,
|
| 354 |
+
"eval_cos_loss": 0.40689416910143206,
|
| 355 |
+
"eval_cov": 0.08788992825164098,
|
| 356 |
+
"eval_cov_loss": 0.01204854997563852,
|
| 357 |
+
"eval_glb_loss": 0.0002671638515636362,
|
| 358 |
+
"eval_global_kurtosis": 2.9975850026901454,
|
| 359 |
+
"eval_global_mean": -0.006530939716182343,
|
| 360 |
+
"eval_global_var": 1.1122132741152968,
|
| 361 |
+
"eval_krt_loss": 0.0002464600221175453,
|
| 362 |
+
"eval_loss": 0.3173055400451024,
|
| 363 |
+
"eval_mean_loss": 5.356442062072988e-05,
|
| 364 |
+
"eval_mse_loss": 0.9764057996610528,
|
| 365 |
+
"eval_per_loss": 5.57458498690526e-06,
|
| 366 |
+
"eval_per_var": 1.0869274400684932,
|
| 367 |
+
"eval_within_var": 1.0837734279022913,
|
| 368 |
+
"eval_wth_loss": 5.189207881040453e-06,
|
| 369 |
+
"step": 5120
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.236478684587317,
|
| 373 |
+
"eval_acr_loss": 0.4548332993978779,
|
| 374 |
+
"eval_across_var": 0.028460351478087304,
|
| 375 |
+
"eval_bleu": 0.9552258623563333,
|
| 376 |
+
"eval_ce_loss": 0.12340653128071463,
|
| 377 |
+
"eval_cos_loss": 0.40689416910143206,
|
| 378 |
+
"eval_cov": 0.08788992825164098,
|
| 379 |
+
"eval_cov_loss": 0.01204854997563852,
|
| 380 |
+
"eval_glb_loss": 0.0002671638515636362,
|
| 381 |
+
"eval_global_kurtosis": 2.9975850026901454,
|
| 382 |
+
"eval_global_mean": -0.006530939716182343,
|
| 383 |
+
"eval_global_var": 1.1122132741152968,
|
| 384 |
+
"eval_krt_loss": 0.0002464600221175453,
|
| 385 |
+
"eval_loss": 0.3173055400451024,
|
| 386 |
+
"eval_mean_loss": 5.356442062072988e-05,
|
| 387 |
+
"eval_mse_loss": 0.9764057996610528,
|
| 388 |
+
"eval_per_loss": 5.57458498690526e-06,
|
| 389 |
+
"eval_per_var": 1.0869274400684932,
|
| 390 |
+
"eval_runtime": 154.1747,
|
| 391 |
+
"eval_samples_per_second": 181.567,
|
| 392 |
+
"eval_steps_per_second": 2.841,
|
| 393 |
+
"eval_within_var": 1.0837734279022913,
|
| 394 |
+
"eval_wth_loss": 5.189207881040453e-06,
|
| 395 |
+
"step": 5120
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.24830261881668284,
|
| 399 |
+
"grad_norm": 0.08900712430477142,
|
| 400 |
+
"learning_rate": 4.401069561246422e-05,
|
| 401 |
+
"loss": 0.3965,
|
| 402 |
+
"step": 5376
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.2601265530460487,
|
| 406 |
+
"grad_norm": 0.08696399629116058,
|
| 407 |
+
"learning_rate": 4.337515714516545e-05,
|
| 408 |
+
"loss": 0.3747,
|
| 409 |
+
"step": 5632
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.27195048727541454,
|
| 413 |
+
"grad_norm": 0.09633397310972214,
|
| 414 |
+
"learning_rate": 4.2712673707468434e-05,
|
| 415 |
+
"loss": 0.3594,
|
| 416 |
+
"step": 5888
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.2837744215047804,
|
| 420 |
+
"grad_norm": 0.08299960196018219,
|
| 421 |
+
"learning_rate": 4.202421675210565e-05,
|
| 422 |
+
"loss": 0.3425,
|
| 423 |
+
"step": 6144
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.2837744215047804,
|
| 427 |
+
"eval_acr_loss": 0.444739139651599,
|
| 428 |
+
"eval_across_var": 0.037298334392476575,
|
| 429 |
+
"eval_bleu": 0.9670970291714691,
|
| 430 |
+
"eval_ce_loss": 0.08759757838122649,
|
| 431 |
+
"eval_cos_loss": 0.34691714662122947,
|
| 432 |
+
"eval_cov": 0.08670684953802797,
|
| 433 |
+
"eval_cov_loss": 0.011868421341288307,
|
| 434 |
+
"eval_glb_loss": 0.0008387476386425913,
|
| 435 |
+
"eval_global_kurtosis": 2.998767622529644,
|
| 436 |
+
"eval_global_mean": -0.007462886234396669,
|
| 437 |
+
"eval_global_var": 1.1261125677796804,
|
| 438 |
+
"eval_krt_loss": 0.00038161360164841305,
|
| 439 |
+
"eval_loss": 0.2573624641111452,
|
| 440 |
+
"eval_mean_loss": 6.83625881451915e-05,
|
| 441 |
+
"eval_mse_loss": 0.8629201449215684,
|
| 442 |
+
"eval_per_loss": 0.00010140475162446005,
|
| 443 |
+
"eval_per_var": 1.1030652468607305,
|
| 444 |
+
"eval_within_var": 1.0889683330440085,
|
| 445 |
+
"eval_wth_loss": 2.5263947968048936e-05,
|
| 446 |
+
"step": 6144
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.2837744215047804,
|
| 450 |
+
"eval_acr_loss": 0.444739139651599,
|
| 451 |
+
"eval_across_var": 0.037298334392476575,
|
| 452 |
+
"eval_bleu": 0.9670970291714691,
|
| 453 |
+
"eval_ce_loss": 0.08759757838122649,
|
| 454 |
+
"eval_cos_loss": 0.34691714662122947,
|
| 455 |
+
"eval_cov": 0.08670684953802797,
|
| 456 |
+
"eval_cov_loss": 0.011868421341288307,
|
| 457 |
+
"eval_glb_loss": 0.0008387476386425913,
|
| 458 |
+
"eval_global_kurtosis": 2.998767622529644,
|
| 459 |
+
"eval_global_mean": -0.007462886234396669,
|
| 460 |
+
"eval_global_var": 1.1261125677796804,
|
| 461 |
+
"eval_krt_loss": 0.00038161360164841305,
|
| 462 |
+
"eval_loss": 0.2573624641111452,
|
| 463 |
+
"eval_mean_loss": 6.83625881451915e-05,
|
| 464 |
+
"eval_mse_loss": 0.8629201449215684,
|
| 465 |
+
"eval_per_loss": 0.00010140475162446005,
|
| 466 |
+
"eval_per_var": 1.1030652468607305,
|
| 467 |
+
"eval_runtime": 153.8249,
|
| 468 |
+
"eval_samples_per_second": 181.98,
|
| 469 |
+
"eval_steps_per_second": 2.847,
|
| 470 |
+
"eval_within_var": 1.0889683330440085,
|
| 471 |
+
"eval_wth_loss": 2.5263947968048936e-05,
|
| 472 |
+
"step": 6144
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.2955983557341462,
|
| 476 |
+
"grad_norm": 0.08428088575601578,
|
| 477 |
+
"learning_rate": 4.131079581886694e-05,
|
| 478 |
+
"loss": 0.325,
|
| 479 |
+
"step": 6400
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.30742228996351206,
|
| 483 |
+
"grad_norm": 0.08620952814817429,
|
| 484 |
+
"learning_rate": 4.057345705423016e-05,
|
| 485 |
+
"loss": 0.3121,
|
| 486 |
+
"step": 6656
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.3192462241928779,
|
| 490 |
+
"grad_norm": 0.08939851075410843,
|
| 491 |
+
"learning_rate": 3.981328167731251e-05,
|
| 492 |
+
"loss": 0.3016,
|
| 493 |
+
"step": 6912
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.3310701584222438,
|
| 497 |
+
"grad_norm": 0.08685296773910522,
|
| 498 |
+
"learning_rate": 3.9031384394391954e-05,
|
| 499 |
+
"loss": 0.2875,
|
| 500 |
+
"step": 7168
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.3310701584222438,
|
| 504 |
+
"eval_acr_loss": 0.4308900043164214,
|
| 505 |
+
"eval_across_var": 0.04932180811230042,
|
| 506 |
+
"eval_bleu": 0.9757814026870458,
|
| 507 |
+
"eval_ce_loss": 0.0650875303026748,
|
| 508 |
+
"eval_cos_loss": 0.3049790498600703,
|
| 509 |
+
"eval_cov": 0.0888557608269121,
|
| 510 |
+
"eval_cov_loss": 0.012648627446904847,
|
| 511 |
+
"eval_glb_loss": 0.0011189478131933299,
|
| 512 |
+
"eval_global_kurtosis": 3.0107862796957634,
|
| 513 |
+
"eval_global_mean": -0.007322731763804885,
|
| 514 |
+
"eval_global_var": 1.130718910530822,
|
| 515 |
+
"eval_krt_loss": 0.0008007638278780262,
|
| 516 |
+
"eval_loss": 0.217820766683855,
|
| 517 |
+
"eval_mean_loss": 6.851153396431181e-05,
|
| 518 |
+
"eval_mse_loss": 0.7866716142658774,
|
| 519 |
+
"eval_per_loss": 0.0001979704842660462,
|
| 520 |
+
"eval_per_var": 1.1079815032819635,
|
| 521 |
+
"eval_within_var": 1.0816122696279935,
|
| 522 |
+
"eval_wth_loss": 1.0462231934273432e-05,
|
| 523 |
+
"step": 7168
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 0.3310701584222438,
|
| 527 |
+
"eval_acr_loss": 0.4308900043164214,
|
| 528 |
+
"eval_across_var": 0.04932180811230042,
|
| 529 |
+
"eval_bleu": 0.9757814026870458,
|
| 530 |
+
"eval_ce_loss": 0.0650875303026748,
|
| 531 |
+
"eval_cos_loss": 0.3049790498600703,
|
| 532 |
+
"eval_cov": 0.0888557608269121,
|
| 533 |
+
"eval_cov_loss": 0.012648627446904847,
|
| 534 |
+
"eval_glb_loss": 0.0011189478131933299,
|
| 535 |
+
"eval_global_kurtosis": 3.0107862796957634,
|
| 536 |
+
"eval_global_mean": -0.007322731763804885,
|
| 537 |
+
"eval_global_var": 1.130718910530822,
|
| 538 |
+
"eval_krt_loss": 0.0008007638278780262,
|
| 539 |
+
"eval_loss": 0.217820766683855,
|
| 540 |
+
"eval_mean_loss": 6.851153396431181e-05,
|
| 541 |
+
"eval_mse_loss": 0.7866716142658774,
|
| 542 |
+
"eval_per_loss": 0.0001979704842660462,
|
| 543 |
+
"eval_per_var": 1.1079815032819635,
|
| 544 |
+
"eval_runtime": 151.7662,
|
| 545 |
+
"eval_samples_per_second": 184.448,
|
| 546 |
+
"eval_steps_per_second": 2.886,
|
| 547 |
+
"eval_within_var": 1.0816122696279935,
|
| 548 |
+
"eval_wth_loss": 1.0462231934273432e-05,
|
| 549 |
+
"step": 7168
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.34289409265160964,
|
| 553 |
+
"grad_norm": 0.08307478576898575,
|
| 554 |
+
"learning_rate": 3.822891176432382e-05,
|
| 555 |
+
"loss": 0.2798,
|
| 556 |
+
"step": 7424
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.3547180268809755,
|
| 560 |
+
"grad_norm": 0.07975339144468307,
|
| 561 |
+
"learning_rate": 3.7407040517249335e-05,
|
| 562 |
+
"loss": 0.2702,
|
| 563 |
+
"step": 7680
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.3665419611103413,
|
| 567 |
+
"grad_norm": 0.08212888985872269,
|
| 568 |
+
"learning_rate": 3.6566975829061614e-05,
|
| 569 |
+
"loss": 0.2607,
|
| 570 |
+
"step": 7936
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.37836589533970716,
|
| 574 |
+
"grad_norm": 0.07644044607877731,
|
| 575 |
+
"learning_rate": 3.5709949554159355e-05,
|
| 576 |
+
"loss": 0.2519,
|
| 577 |
+
"step": 8192
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.37836589533970716,
|
| 581 |
+
"eval_acr_loss": 0.41207800308863324,
|
| 582 |
+
"eval_across_var": 0.0661766531167033,
|
| 583 |
+
"eval_bleu": 0.9805728687340939,
|
| 584 |
+
"eval_ce_loss": 0.05073687865411582,
|
| 585 |
+
"eval_cos_loss": 0.2765354519905565,
|
| 586 |
+
"eval_cov": 0.09245990073844178,
|
| 587 |
+
"eval_cov_loss": 0.013950884125310264,
|
| 588 |
+
"eval_glb_loss": 0.0015832305986751307,
|
| 589 |
+
"eval_global_kurtosis": 3.0033102030079113,
|
| 590 |
+
"eval_global_mean": -0.008325645232309489,
|
| 591 |
+
"eval_global_var": 1.1371356842180365,
|
| 592 |
+
"eval_krt_loss": 0.0005490112146340372,
|
| 593 |
+
"eval_loss": 0.19166131559164012,
|
| 594 |
+
"eval_mean_loss": 8.809174024773519e-05,
|
| 595 |
+
"eval_mse_loss": 0.7389907115670644,
|
| 596 |
+
"eval_per_loss": 0.00044034952490991304,
|
| 597 |
+
"eval_per_var": 1.1152143086472603,
|
| 598 |
+
"eval_within_var": 1.0713012561406174,
|
| 599 |
+
"eval_wth_loss": 1.391253488354359e-06,
|
| 600 |
+
"step": 8192
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.37836589533970716,
|
| 604 |
+
"eval_acr_loss": 0.41207800308863324,
|
| 605 |
+
"eval_across_var": 0.0661766531167033,
|
| 606 |
+
"eval_bleu": 0.9805728687340939,
|
| 607 |
+
"eval_ce_loss": 0.05073687865411582,
|
| 608 |
+
"eval_cos_loss": 0.2765354519905565,
|
| 609 |
+
"eval_cov": 0.09245990073844178,
|
| 610 |
+
"eval_cov_loss": 0.013950884125310264,
|
| 611 |
+
"eval_glb_loss": 0.0015832305986751307,
|
| 612 |
+
"eval_global_kurtosis": 3.0033102030079113,
|
| 613 |
+
"eval_global_mean": -0.008325645232309489,
|
| 614 |
+
"eval_global_var": 1.1371356842180365,
|
| 615 |
+
"eval_krt_loss": 0.0005490112146340372,
|
| 616 |
+
"eval_loss": 0.19166131559164012,
|
| 617 |
+
"eval_mean_loss": 8.809174024773519e-05,
|
| 618 |
+
"eval_mse_loss": 0.7389907115670644,
|
| 619 |
+
"eval_per_loss": 0.00044034952490991304,
|
| 620 |
+
"eval_per_var": 1.1152143086472603,
|
| 621 |
+
"eval_runtime": 151.0158,
|
| 622 |
+
"eval_samples_per_second": 185.365,
|
| 623 |
+
"eval_steps_per_second": 2.9,
|
| 624 |
+
"eval_within_var": 1.0713012561406174,
|
| 625 |
+
"eval_wth_loss": 1.391253488354359e-06,
|
| 626 |
+
"step": 8192
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.390189829569073,
|
| 630 |
+
"grad_norm": 0.07752202451229095,
|
| 631 |
+
"learning_rate": 3.483721841907964e-05,
|
| 632 |
+
"loss": 0.2462,
|
| 633 |
+
"step": 8448
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.4020137637984389,
|
| 637 |
+
"grad_norm": 0.08402163535356522,
|
| 638 |
+
"learning_rate": 3.3953554020219734e-05,
|
| 639 |
+
"loss": 0.2397,
|
| 640 |
+
"step": 8704
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.41383769802780473,
|
| 644 |
+
"grad_norm": 0.08829676359891891,
|
| 645 |
+
"learning_rate": 3.305332229492516e-05,
|
| 646 |
+
"loss": 0.2323,
|
| 647 |
+
"step": 8960
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.4256616322571706,
|
| 651 |
+
"grad_norm": 0.07489398866891861,
|
| 652 |
+
"learning_rate": 3.214128133561262e-05,
|
| 653 |
+
"loss": 0.2275,
|
| 654 |
+
"step": 9216
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.4256616322571706,
|
| 658 |
+
"eval_acr_loss": 0.3899558907094067,
|
| 659 |
+
"eval_across_var": 0.08639241476037186,
|
| 660 |
+
"eval_bleu": 0.9839023939205267,
|
| 661 |
+
"eval_ce_loss": 0.04116568632730066,
|
| 662 |
+
"eval_cos_loss": 0.25615145446366916,
|
| 663 |
+
"eval_cov": 0.09534583244149543,
|
| 664 |
+
"eval_cov_loss": 0.015230727917833687,
|
| 665 |
+
"eval_glb_loss": 0.001839591442792664,
|
| 666 |
+
"eval_global_kurtosis": 3.0162572811727655,
|
| 667 |
+
"eval_global_mean": -0.008430217362974333,
|
| 668 |
+
"eval_global_var": 1.1393898045091324,
|
| 669 |
+
"eval_krt_loss": 0.001037847468692951,
|
| 670 |
+
"eval_loss": 0.17338106053196675,
|
| 671 |
+
"eval_mean_loss": 9.661306929678656e-05,
|
| 672 |
+
"eval_mse_loss": 0.7077833269829075,
|
| 673 |
+
"eval_per_loss": 0.0006317324198246723,
|
| 674 |
+
"eval_per_var": 1.1172164847317352,
|
| 675 |
+
"eval_within_var": 1.053544073616533,
|
| 676 |
+
"eval_wth_loss": 0.0,
|
| 677 |
+
"step": 9216
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.4256616322571706,
|
| 681 |
+
"eval_acr_loss": 0.3899558907094067,
|
| 682 |
+
"eval_across_var": 0.08639241476037186,
|
| 683 |
+
"eval_bleu": 0.9839023939205267,
|
| 684 |
+
"eval_ce_loss": 0.04116568632730066,
|
| 685 |
+
"eval_cos_loss": 0.25615145446366916,
|
| 686 |
+
"eval_cov": 0.09534583244149543,
|
| 687 |
+
"eval_cov_loss": 0.015230727917833687,
|
| 688 |
+
"eval_glb_loss": 0.001839591442792664,
|
| 689 |
+
"eval_global_kurtosis": 3.0162572811727655,
|
| 690 |
+
"eval_global_mean": -0.008430217362974333,
|
| 691 |
+
"eval_global_var": 1.1393898045091324,
|
| 692 |
+
"eval_krt_loss": 0.001037847468692951,
|
| 693 |
+
"eval_loss": 0.17338106053196675,
|
| 694 |
+
"eval_mean_loss": 9.661306929678656e-05,
|
| 695 |
+
"eval_mse_loss": 0.7077833269829075,
|
| 696 |
+
"eval_per_loss": 0.0006317324198246723,
|
| 697 |
+
"eval_per_var": 1.1172164847317352,
|
| 698 |
+
"eval_runtime": 149.7127,
|
| 699 |
+
"eval_samples_per_second": 186.978,
|
| 700 |
+
"eval_steps_per_second": 2.926,
|
| 701 |
+
"eval_within_var": 1.053544073616533,
|
| 702 |
+
"eval_wth_loss": 0.0,
|
| 703 |
+
"step": 9216
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.4374855664865364,
|
| 707 |
+
"grad_norm": 0.07958604395389557,
|
| 708 |
+
"learning_rate": 3.1218768541274476e-05,
|
| 709 |
+
"loss": 0.2225,
|
| 710 |
+
"step": 9472
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.44930950071590225,
|
| 714 |
+
"grad_norm": 0.08125226199626923,
|
| 715 |
+
"learning_rate": 3.028713666659683e-05,
|
| 716 |
+
"loss": 0.219,
|
| 717 |
+
"step": 9728
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 0.4611334349452681,
|
| 721 |
+
"grad_norm": 0.08044584840536118,
|
| 722 |
+
"learning_rate": 2.9347751838306454e-05,
|
| 723 |
+
"loss": 0.2136,
|
| 724 |
+
"step": 9984
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 0.472957369174634,
|
| 728 |
+
"grad_norm": 0.09431233257055283,
|
| 729 |
+
"learning_rate": 2.840199155190943e-05,
|
| 730 |
+
"loss": 0.2109,
|
| 731 |
+
"step": 10240
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 0.472957369174634,
|
| 735 |
+
"eval_acr_loss": 0.371521062094327,
|
| 736 |
+
"eval_across_var": 0.10420215363746092,
|
| 737 |
+
"eval_bleu": 0.9864977390539889,
|
| 738 |
+
"eval_ce_loss": 0.03421062290813076,
|
| 739 |
+
"eval_cos_loss": 0.24110814265600622,
|
| 740 |
+
"eval_cov": 0.09938139893692922,
|
| 741 |
+
"eval_cov_loss": 0.016782419545550462,
|
| 742 |
+
"eval_glb_loss": 0.0025077612427728364,
|
| 743 |
+
"eval_global_kurtosis": 3.000300804229632,
|
| 744 |
+
"eval_global_mean": -0.008482733817949686,
|
| 745 |
+
"eval_global_var": 1.1459448130707763,
|
| 746 |
+
"eval_krt_loss": 0.0005616036675252787,
|
| 747 |
+
"eval_loss": 0.16001523779407484,
|
| 748 |
+
"eval_mean_loss": 0.00010505710818644507,
|
| 749 |
+
"eval_mse_loss": 0.6870357463621113,
|
| 750 |
+
"eval_per_loss": 0.0010411720005544605,
|
| 751 |
+
"eval_per_var": 1.1233344927226028,
|
| 752 |
+
"eval_within_var": 1.042447058849683,
|
| 753 |
+
"eval_wth_loss": 0.0,
|
| 754 |
+
"step": 10240
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 0.472957369174634,
|
| 758 |
+
"eval_acr_loss": 0.371521062094327,
|
| 759 |
+
"eval_across_var": 0.10420215363746092,
|
| 760 |
+
"eval_bleu": 0.9864977390539889,
|
| 761 |
+
"eval_ce_loss": 0.03421062290813076,
|
| 762 |
+
"eval_cos_loss": 0.24110814265600622,
|
| 763 |
+
"eval_cov": 0.09938139893692922,
|
| 764 |
+
"eval_cov_loss": 0.016782419545550462,
|
| 765 |
+
"eval_glb_loss": 0.0025077612427728364,
|
| 766 |
+
"eval_global_kurtosis": 3.000300804229632,
|
| 767 |
+
"eval_global_mean": -0.008482733817949686,
|
| 768 |
+
"eval_global_var": 1.1459448130707763,
|
| 769 |
+
"eval_krt_loss": 0.0005616036675252787,
|
| 770 |
+
"eval_loss": 0.16001523779407484,
|
| 771 |
+
"eval_mean_loss": 0.00010505710818644507,
|
| 772 |
+
"eval_mse_loss": 0.6870357463621113,
|
| 773 |
+
"eval_per_loss": 0.0010411720005544605,
|
| 774 |
+
"eval_per_var": 1.1233344927226028,
|
| 775 |
+
"eval_runtime": 150.7627,
|
| 776 |
+
"eval_samples_per_second": 185.676,
|
| 777 |
+
"eval_steps_per_second": 2.905,
|
| 778 |
+
"eval_within_var": 1.042447058849683,
|
| 779 |
+
"eval_wth_loss": 0.0,
|
| 780 |
+
"step": 10240
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 0.48478130340399983,
|
| 784 |
+
"grad_norm": 0.07165589183568954,
|
| 785 |
+
"learning_rate": 2.745124265175868e-05,
|
| 786 |
+
"loss": 0.2062,
|
| 787 |
+
"step": 10496
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 0.49660523763336567,
|
| 791 |
+
"grad_norm": 0.07731898874044418,
|
| 792 |
+
"learning_rate": 2.6496899297412598e-05,
|
| 793 |
+
"loss": 0.2023,
|
| 794 |
+
"step": 10752
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.5084291718627315,
|
| 798 |
+
"grad_norm": 0.08252639323472977,
|
| 799 |
+
"learning_rate": 2.5544099852682395e-05,
|
| 800 |
+
"loss": 0.2011,
|
| 801 |
+
"step": 11008
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.5202531060920974,
|
| 805 |
+
"grad_norm": 0.09438033401966095,
|
| 806 |
+
"learning_rate": 2.4586769464065254e-05,
|
| 807 |
+
"loss": 0.1971,
|
| 808 |
+
"step": 11264
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 0.5202531060920974,
|
| 812 |
+
"eval_acr_loss": 0.3526782588735563,
|
| 813 |
+
"eval_across_var": 0.1233543821240533,
|
| 814 |
+
"eval_bleu": 0.9885170657049208,
|
| 815 |
+
"eval_ce_loss": 0.02941183759256789,
|
| 816 |
+
"eval_cos_loss": 0.2296683646432341,
|
| 817 |
+
"eval_cov": 0.10448926115689212,
|
| 818 |
+
"eval_cov_loss": 0.018818640007667208,
|
| 819 |
+
"eval_glb_loss": 0.0035778092502643057,
|
| 820 |
+
"eval_global_kurtosis": 3.010114621898355,
|
| 821 |
+
"eval_global_mean": -0.011101974472063318,
|
| 822 |
+
"eval_global_var": 1.1552310751997716,
|
| 823 |
+
"eval_krt_loss": 0.0011141615731220227,
|
| 824 |
+
"eval_loss": 0.15028831549839342,
|
| 825 |
+
"eval_mean_loss": 0.0001692649679361995,
|
| 826 |
+
"eval_mse_loss": 0.6726388316176254,
|
| 827 |
+
"eval_per_loss": 0.0018812230319723378,
|
| 828 |
+
"eval_per_var": 1.1340878281963471,
|
| 829 |
+
"eval_within_var": 1.0327191478041209,
|
| 830 |
+
"eval_wth_loss": 0.0,
|
| 831 |
+
"step": 11264
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.5202531060920974,
|
| 835 |
+
"eval_acr_loss": 0.3526782588735563,
|
| 836 |
+
"eval_across_var": 0.1233543821240533,
|
| 837 |
+
"eval_bleu": 0.9885170657049208,
|
| 838 |
+
"eval_ce_loss": 0.02941183759256789,
|
| 839 |
+
"eval_cos_loss": 0.2296683646432341,
|
| 840 |
+
"eval_cov": 0.10448926115689212,
|
| 841 |
+
"eval_cov_loss": 0.018818640007667208,
|
| 842 |
+
"eval_glb_loss": 0.0035778092502643057,
|
| 843 |
+
"eval_global_kurtosis": 3.010114621898355,
|
| 844 |
+
"eval_global_mean": -0.011101974472063318,
|
| 845 |
+
"eval_global_var": 1.1552310751997716,
|
| 846 |
+
"eval_krt_loss": 0.0011141615731220227,
|
| 847 |
+
"eval_loss": 0.15028831549839342,
|
| 848 |
+
"eval_mean_loss": 0.0001692649679361995,
|
| 849 |
+
"eval_mse_loss": 0.6726388316176254,
|
| 850 |
+
"eval_per_loss": 0.0018812230319723378,
|
| 851 |
+
"eval_per_var": 1.1340878281963471,
|
| 852 |
+
"eval_runtime": 150.1371,
|
| 853 |
+
"eval_samples_per_second": 186.45,
|
| 854 |
+
"eval_steps_per_second": 2.917,
|
| 855 |
+
"eval_within_var": 1.0327191478041209,
|
| 856 |
+
"eval_wth_loss": 0.0,
|
| 857 |
+
"step": 11264
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.5320770403214632,
|
| 861 |
+
"grad_norm": 0.08242896944284439,
|
| 862 |
+
"learning_rate": 2.3630045028609248e-05,
|
| 863 |
+
"loss": 0.1941,
|
| 864 |
+
"step": 11520
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.5439009745508291,
|
| 868 |
+
"grad_norm": 0.12015263736248016,
|
| 869 |
+
"learning_rate": 2.267532946828065e-05,
|
| 870 |
+
"loss": 0.1923,
|
| 871 |
+
"step": 11776
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.5557249087801949,
|
| 875 |
+
"grad_norm": 0.07551924884319305,
|
| 876 |
+
"learning_rate": 2.1724022759270597e-05,
|
| 877 |
+
"loss": 0.19,
|
| 878 |
+
"step": 12032
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.5675488430095608,
|
| 882 |
+
"grad_norm": 0.09451667219400406,
|
| 883 |
+
"learning_rate": 2.0777519879097458e-05,
|
| 884 |
+
"loss": 0.1889,
|
| 885 |
+
"step": 12288
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 0.5675488430095608,
|
| 889 |
+
"eval_acr_loss": 0.34067982850281614,
|
| 890 |
+
"eval_across_var": 0.1346761514964305,
|
| 891 |
+
"eval_bleu": 0.9895254511324281,
|
| 892 |
+
"eval_ce_loss": 0.0263222230802456,
|
| 893 |
+
"eval_cos_loss": 0.22149913948556604,
|
| 894 |
+
"eval_cov": 0.10440648745184075,
|
| 895 |
+
"eval_cov_loss": 0.019090854430987955,
|
| 896 |
+
"eval_glb_loss": 0.0031285923291081673,
|
| 897 |
+
"eval_global_kurtosis": 3.0161606329216806,
|
| 898 |
+
"eval_global_mean": -0.01319564518318873,
|
| 899 |
+
"eval_global_var": 1.1500695633561644,
|
| 900 |
+
"eval_krt_loss": 0.0014608212088625325,
|
| 901 |
+
"eval_loss": 0.14371125602490825,
|
| 902 |
+
"eval_mean_loss": 0.0002307435466704484,
|
| 903 |
+
"eval_mse_loss": 0.6642545581408287,
|
| 904 |
+
"eval_per_loss": 0.00148715251082838,
|
| 905 |
+
"eval_per_var": 1.1267546910673516,
|
| 906 |
+
"eval_within_var": 1.0164183696655378,
|
| 907 |
+
"eval_wth_loss": 0.0,
|
| 908 |
+
"step": 12288
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.5675488430095608,
|
| 912 |
+
"eval_acr_loss": 0.34067982850281614,
|
| 913 |
+
"eval_across_var": 0.1346761514964305,
|
| 914 |
+
"eval_bleu": 0.9895254511324281,
|
| 915 |
+
"eval_ce_loss": 0.0263222230802456,
|
| 916 |
+
"eval_cos_loss": 0.22149913948556604,
|
| 917 |
+
"eval_cov": 0.10440648745184075,
|
| 918 |
+
"eval_cov_loss": 0.019090854430987955,
|
| 919 |
+
"eval_glb_loss": 0.0031285923291081673,
|
| 920 |
+
"eval_global_kurtosis": 3.0161606329216806,
|
| 921 |
+
"eval_global_mean": -0.01319564518318873,
|
| 922 |
+
"eval_global_var": 1.1500695633561644,
|
| 923 |
+
"eval_krt_loss": 0.0014608212088625325,
|
| 924 |
+
"eval_loss": 0.14371125602490825,
|
| 925 |
+
"eval_mean_loss": 0.0002307435466704484,
|
| 926 |
+
"eval_mse_loss": 0.6642545581408287,
|
| 927 |
+
"eval_per_loss": 0.00148715251082838,
|
| 928 |
+
"eval_per_var": 1.1267546910673516,
|
| 929 |
+
"eval_runtime": 149.6169,
|
| 930 |
+
"eval_samples_per_second": 187.098,
|
| 931 |
+
"eval_steps_per_second": 2.927,
|
| 932 |
+
"eval_within_var": 1.0164183696655378,
|
| 933 |
+
"eval_wth_loss": 0.0,
|
| 934 |
+
"step": 12288
|
| 935 |
+
}
|
| 936 |
+
],
|
| 937 |
+
"logging_steps": 256,
|
| 938 |
+
"max_steps": 21651,
|
| 939 |
+
"num_input_tokens_seen": 0,
|
| 940 |
+
"num_train_epochs": 1,
|
| 941 |
+
"save_steps": 1024,
|
| 942 |
+
"stateful_callbacks": {
|
| 943 |
+
"TrainerControl": {
|
| 944 |
+
"args": {
|
| 945 |
+
"should_epoch_stop": false,
|
| 946 |
+
"should_evaluate": false,
|
| 947 |
+
"should_log": false,
|
| 948 |
+
"should_save": true,
|
| 949 |
+
"should_training_stop": false
|
| 950 |
+
},
|
| 951 |
+
"attributes": {}
|
| 952 |
+
}
|
| 953 |
+
},
|
| 954 |
+
"total_flos": 0.0,
|
| 955 |
+
"train_batch_size": 64,
|
| 956 |
+
"trial_name": null,
|
| 957 |
+
"trial_params": null
|
| 958 |
+
}
|
checkpoints-v2.6/checkpoint-12288/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae70ca1441ea90d34c54e8c897a025ccc9b6d942bad37cfab215bb440b2ecd4d
|
| 3 |
+
size 5777
|