checkpoint-1907
Browse files- gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/README.md +0 -0
- gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/adapter_config.json +0 -0
- gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/adapter_model.safetensors +1 -1
- gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/optimizer.pt +1 -1
- gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/rng_state.pth +1 -1
- gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/scheduler.pt +1 -1
- gemma-3-1b-it/checkpoint-1907/trainer_state.json +1420 -0
- gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/training_args.bin +0 -0
- gemma-3-1b-it/checkpoint-500/trainer_state.json +0 -392
- gemma-3-1b-it/tb/events.out.tfevents.1789955877.fed349279e2c.8616.0 +2 -2
- gemma-3-1b-it/test_eval/metrics.json +2 -2
- gemma-3-1b-it/test_eval/predictions.jsonl +0 -0
gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/README.md
RENAMED
|
File without changes
|
gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/adapter_config.json
RENAMED
|
File without changes
|
gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/adapter_model.safetensors
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 104414464
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9fe6c26ed636257777551f7b24e54c44239486928672902cdb026cac31856d20
|
| 3 |
size 104414464
|
gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/optimizer.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 209043559
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:14fd928b01c924bc4d81ecd11b456728bbbb1b3a70a1d229614c038dc838057d
|
| 3 |
size 209043559
|
gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/rng_state.pth
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1dc646354c0f9731caf276286a8b6fb3e3c9e56a3282daacf150d5ef7fbebc7e
|
| 3 |
size 14645
|
gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/scheduler.pt
RENAMED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69415cb16c68d4576bac93195b508d63936b7b2aaa5cc9f7af5bcef6ca3ab038
|
| 3 |
size 1465
|
gemma-3-1b-it/checkpoint-1907/trainer_state.json
ADDED
|
@@ -0,0 +1,1420 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1907,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.005245901639344262,
|
| 14 |
+
"grad_norm": 5.850006580352783,
|
| 15 |
+
"learning_rate": 1.8750000000000002e-05,
|
| 16 |
+
"loss": 1.626298713684082,
|
| 17 |
+
"step": 10
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.010491803278688525,
|
| 21 |
+
"grad_norm": 1.7961169481277466,
|
| 22 |
+
"learning_rate": 3.958333333333333e-05,
|
| 23 |
+
"loss": 1.3221087455749512,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.015737704918032787,
|
| 28 |
+
"grad_norm": 1.4881651401519775,
|
| 29 |
+
"learning_rate": 6.041666666666667e-05,
|
| 30 |
+
"loss": 1.1436286926269532,
|
| 31 |
+
"step": 30
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.02098360655737705,
|
| 35 |
+
"grad_norm": 1.0592900514602661,
|
| 36 |
+
"learning_rate": 8.125000000000001e-05,
|
| 37 |
+
"loss": 0.9118364334106446,
|
| 38 |
+
"step": 40
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.02622950819672131,
|
| 42 |
+
"grad_norm": 1.3136152029037476,
|
| 43 |
+
"learning_rate": 0.00010208333333333333,
|
| 44 |
+
"loss": 0.9902483940124511,
|
| 45 |
+
"step": 50
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.031475409836065574,
|
| 49 |
+
"grad_norm": 1.0517693758010864,
|
| 50 |
+
"learning_rate": 0.00012291666666666668,
|
| 51 |
+
"loss": 0.9533065795898438,
|
| 52 |
+
"step": 60
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.036721311475409836,
|
| 56 |
+
"grad_norm": 1.0763620138168335,
|
| 57 |
+
"learning_rate": 0.00014375,
|
| 58 |
+
"loss": 0.9590832710266113,
|
| 59 |
+
"step": 70
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.0419672131147541,
|
| 63 |
+
"grad_norm": 1.126420497894287,
|
| 64 |
+
"learning_rate": 0.00016458333333333334,
|
| 65 |
+
"loss": 0.8696105003356933,
|
| 66 |
+
"step": 80
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.04721311475409836,
|
| 70 |
+
"grad_norm": 1.0479974746704102,
|
| 71 |
+
"learning_rate": 0.00018541666666666668,
|
| 72 |
+
"loss": 0.8585530281066894,
|
| 73 |
+
"step": 90
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.05245901639344262,
|
| 77 |
+
"grad_norm": 0.9453676342964172,
|
| 78 |
+
"learning_rate": 0.0001996686913307565,
|
| 79 |
+
"loss": 0.81344633102417,
|
| 80 |
+
"step": 100
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.057704918032786885,
|
| 84 |
+
"grad_norm": 0.9837343692779541,
|
| 85 |
+
"learning_rate": 0.0001985643290999448,
|
| 86 |
+
"loss": 0.9181200981140136,
|
| 87 |
+
"step": 110
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.06295081967213115,
|
| 91 |
+
"grad_norm": 0.8958956003189087,
|
| 92 |
+
"learning_rate": 0.00019745996686913308,
|
| 93 |
+
"loss": 0.826357650756836,
|
| 94 |
+
"step": 120
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.06819672131147542,
|
| 98 |
+
"grad_norm": 0.9927870631217957,
|
| 99 |
+
"learning_rate": 0.0001963556046383214,
|
| 100 |
+
"loss": 0.8804688453674316,
|
| 101 |
+
"step": 130
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.07344262295081967,
|
| 105 |
+
"grad_norm": 0.8682142496109009,
|
| 106 |
+
"learning_rate": 0.00019525124240750967,
|
| 107 |
+
"loss": 0.8082976341247559,
|
| 108 |
+
"step": 140
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.07868852459016394,
|
| 112 |
+
"grad_norm": 0.837242841720581,
|
| 113 |
+
"learning_rate": 0.00019414688017669799,
|
| 114 |
+
"loss": 0.8015812873840332,
|
| 115 |
+
"step": 150
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.0839344262295082,
|
| 119 |
+
"grad_norm": 1.099856972694397,
|
| 120 |
+
"learning_rate": 0.00019304251794588625,
|
| 121 |
+
"loss": 0.7791732311248779,
|
| 122 |
+
"step": 160
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.08918032786885247,
|
| 126 |
+
"grad_norm": 0.9170840978622437,
|
| 127 |
+
"learning_rate": 0.00019193815571507457,
|
| 128 |
+
"loss": 0.8227507591247558,
|
| 129 |
+
"step": 170
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.09442622950819672,
|
| 133 |
+
"grad_norm": 0.8296833038330078,
|
| 134 |
+
"learning_rate": 0.00019083379348426284,
|
| 135 |
+
"loss": 0.7757059097290039,
|
| 136 |
+
"step": 180
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.09967213114754099,
|
| 140 |
+
"grad_norm": 0.7342070937156677,
|
| 141 |
+
"learning_rate": 0.00018972943125345116,
|
| 142 |
+
"loss": 0.7799826622009277,
|
| 143 |
+
"step": 190
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.10491803278688525,
|
| 147 |
+
"grad_norm": 0.7804243564605713,
|
| 148 |
+
"learning_rate": 0.00018862506902263943,
|
| 149 |
+
"loss": 0.7913103580474854,
|
| 150 |
+
"step": 200
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.11016393442622952,
|
| 154 |
+
"grad_norm": 0.7638176083564758,
|
| 155 |
+
"learning_rate": 0.00018752070679182775,
|
| 156 |
+
"loss": 0.7876776695251465,
|
| 157 |
+
"step": 210
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.11540983606557377,
|
| 161 |
+
"grad_norm": 0.6659265160560608,
|
| 162 |
+
"learning_rate": 0.00018641634456101602,
|
| 163 |
+
"loss": 0.7114231586456299,
|
| 164 |
+
"step": 220
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.12065573770491804,
|
| 168 |
+
"grad_norm": 0.6473881602287292,
|
| 169 |
+
"learning_rate": 0.00018531198233020434,
|
| 170 |
+
"loss": 0.7239602088928223,
|
| 171 |
+
"step": 230
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.1259016393442623,
|
| 175 |
+
"grad_norm": 0.8192205429077148,
|
| 176 |
+
"learning_rate": 0.0001842076200993926,
|
| 177 |
+
"loss": 0.7307174205780029,
|
| 178 |
+
"step": 240
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.13114754098360656,
|
| 182 |
+
"grad_norm": 0.6621026396751404,
|
| 183 |
+
"learning_rate": 0.00018310325786858092,
|
| 184 |
+
"loss": 0.7196334362030029,
|
| 185 |
+
"step": 250
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.13639344262295083,
|
| 189 |
+
"grad_norm": 0.6377759575843811,
|
| 190 |
+
"learning_rate": 0.0001819988956377692,
|
| 191 |
+
"loss": 0.761665153503418,
|
| 192 |
+
"step": 260
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.14163934426229507,
|
| 196 |
+
"grad_norm": 0.7564328908920288,
|
| 197 |
+
"learning_rate": 0.0001808945334069575,
|
| 198 |
+
"loss": 0.7283723831176758,
|
| 199 |
+
"step": 270
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.14688524590163934,
|
| 203 |
+
"grad_norm": 0.7781689167022705,
|
| 204 |
+
"learning_rate": 0.00017979017117614578,
|
| 205 |
+
"loss": 0.7663159370422363,
|
| 206 |
+
"step": 280
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.1521311475409836,
|
| 210 |
+
"grad_norm": 0.7704823017120361,
|
| 211 |
+
"learning_rate": 0.0001786858089453341,
|
| 212 |
+
"loss": 0.7273806095123291,
|
| 213 |
+
"step": 290
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.15737704918032788,
|
| 217 |
+
"grad_norm": 0.6168962717056274,
|
| 218 |
+
"learning_rate": 0.00017758144671452237,
|
| 219 |
+
"loss": 0.7601676940917969,
|
| 220 |
+
"step": 300
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.16262295081967212,
|
| 224 |
+
"grad_norm": 0.747313380241394,
|
| 225 |
+
"learning_rate": 0.0001764770844837107,
|
| 226 |
+
"loss": 0.7394940853118896,
|
| 227 |
+
"step": 310
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.1678688524590164,
|
| 231 |
+
"grad_norm": 0.6736217737197876,
|
| 232 |
+
"learning_rate": 0.00017537272225289895,
|
| 233 |
+
"loss": 0.674213457107544,
|
| 234 |
+
"step": 320
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.17311475409836066,
|
| 238 |
+
"grad_norm": 0.7672621011734009,
|
| 239 |
+
"learning_rate": 0.00017426836002208728,
|
| 240 |
+
"loss": 0.6635504722595215,
|
| 241 |
+
"step": 330
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.17836065573770493,
|
| 245 |
+
"grad_norm": 0.6182668209075928,
|
| 246 |
+
"learning_rate": 0.00017316399779127554,
|
| 247 |
+
"loss": 0.6947505950927735,
|
| 248 |
+
"step": 340
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.18360655737704917,
|
| 252 |
+
"grad_norm": 0.6231220960617065,
|
| 253 |
+
"learning_rate": 0.00017205963556046384,
|
| 254 |
+
"loss": 0.7503675937652587,
|
| 255 |
+
"step": 350
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.18885245901639344,
|
| 259 |
+
"grad_norm": 0.7776580452919006,
|
| 260 |
+
"learning_rate": 0.00017095527332965213,
|
| 261 |
+
"loss": 0.7351272106170654,
|
| 262 |
+
"step": 360
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.1940983606557377,
|
| 266 |
+
"grad_norm": 0.7008368968963623,
|
| 267 |
+
"learning_rate": 0.00016985091109884042,
|
| 268 |
+
"loss": 0.6939175605773926,
|
| 269 |
+
"step": 370
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.19934426229508198,
|
| 273 |
+
"grad_norm": 0.7504845261573792,
|
| 274 |
+
"learning_rate": 0.00016874654886802872,
|
| 275 |
+
"loss": 0.7262158393859863,
|
| 276 |
+
"step": 380
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.20459016393442622,
|
| 280 |
+
"grad_norm": 0.6740432977676392,
|
| 281 |
+
"learning_rate": 0.000167642186637217,
|
| 282 |
+
"loss": 0.7514342308044434,
|
| 283 |
+
"step": 390
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.2098360655737705,
|
| 287 |
+
"grad_norm": 0.6686267256736755,
|
| 288 |
+
"learning_rate": 0.0001665378244064053,
|
| 289 |
+
"loss": 0.6774350166320801,
|
| 290 |
+
"step": 400
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.21508196721311476,
|
| 294 |
+
"grad_norm": 0.5805259943008423,
|
| 295 |
+
"learning_rate": 0.0001654334621755936,
|
| 296 |
+
"loss": 0.6829588413238525,
|
| 297 |
+
"step": 410
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.22032786885245903,
|
| 301 |
+
"grad_norm": 0.6207637786865234,
|
| 302 |
+
"learning_rate": 0.0001643290999447819,
|
| 303 |
+
"loss": 0.6976052761077881,
|
| 304 |
+
"step": 420
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.22557377049180327,
|
| 308 |
+
"grad_norm": 1.0807863473892212,
|
| 309 |
+
"learning_rate": 0.0001632247377139702,
|
| 310 |
+
"loss": 0.7167800426483154,
|
| 311 |
+
"step": 430
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.23081967213114754,
|
| 315 |
+
"grad_norm": 0.6717860698699951,
|
| 316 |
+
"learning_rate": 0.00016212037548315848,
|
| 317 |
+
"loss": 0.701947021484375,
|
| 318 |
+
"step": 440
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.2360655737704918,
|
| 322 |
+
"grad_norm": 0.6229609847068787,
|
| 323 |
+
"learning_rate": 0.00016101601325234678,
|
| 324 |
+
"loss": 0.6681315898895264,
|
| 325 |
+
"step": 450
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.24131147540983608,
|
| 329 |
+
"grad_norm": 0.8894275426864624,
|
| 330 |
+
"learning_rate": 0.00015991165102153507,
|
| 331 |
+
"loss": 0.7274843692779541,
|
| 332 |
+
"step": 460
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.24655737704918032,
|
| 336 |
+
"grad_norm": 0.8598907589912415,
|
| 337 |
+
"learning_rate": 0.00015880728879072336,
|
| 338 |
+
"loss": 0.6865033149719239,
|
| 339 |
+
"step": 470
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.2518032786885246,
|
| 343 |
+
"grad_norm": 1.4645612239837646,
|
| 344 |
+
"learning_rate": 0.00015770292655991166,
|
| 345 |
+
"loss": 0.6898164749145508,
|
| 346 |
+
"step": 480
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.25704918032786883,
|
| 350 |
+
"grad_norm": 0.7187657952308655,
|
| 351 |
+
"learning_rate": 0.00015659856432909995,
|
| 352 |
+
"loss": 0.7331682682037354,
|
| 353 |
+
"step": 490
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.26229508196721313,
|
| 357 |
+
"grad_norm": 0.6352348327636719,
|
| 358 |
+
"learning_rate": 0.00015549420209828824,
|
| 359 |
+
"loss": 0.6853579998016357,
|
| 360 |
+
"step": 500
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.26229508196721313,
|
| 364 |
+
"eval_loss": 0.6940423250198364,
|
| 365 |
+
"eval_runtime": 327.5265,
|
| 366 |
+
"eval_samples_per_second": 34.443,
|
| 367 |
+
"eval_steps_per_second": 8.613,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.26229508196721313,
|
| 372 |
+
"step": 500,
|
| 373 |
+
"test_best_wer": 1.1390086753589366,
|
| 374 |
+
"test_exact_match": 0.0,
|
| 375 |
+
"test_hallucination_rate": 0.28125,
|
| 376 |
+
"test_wer": 1.1390086753589366
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 0.26754098360655737,
|
| 380 |
+
"grad_norm": 0.6836485266685486,
|
| 381 |
+
"learning_rate": 0.00015438983986747654,
|
| 382 |
+
"loss": 0.7229701519012451,
|
| 383 |
+
"step": 510
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 0.27278688524590167,
|
| 387 |
+
"grad_norm": 0.6233025193214417,
|
| 388 |
+
"learning_rate": 0.00015328547763666483,
|
| 389 |
+
"loss": 0.7291789531707764,
|
| 390 |
+
"step": 520
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"epoch": 0.2780327868852459,
|
| 394 |
+
"grad_norm": 0.7435803413391113,
|
| 395 |
+
"learning_rate": 0.00015218111540585313,
|
| 396 |
+
"loss": 0.6764275074005127,
|
| 397 |
+
"step": 530
|
| 398 |
+
},
|
| 399 |
+
{
|
| 400 |
+
"epoch": 0.28327868852459015,
|
| 401 |
+
"grad_norm": 0.6755605936050415,
|
| 402 |
+
"learning_rate": 0.00015107675317504142,
|
| 403 |
+
"loss": 0.713279390335083,
|
| 404 |
+
"step": 540
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.28852459016393445,
|
| 408 |
+
"grad_norm": 1.268239974975586,
|
| 409 |
+
"learning_rate": 0.00014997239094422971,
|
| 410 |
+
"loss": 0.7585635662078858,
|
| 411 |
+
"step": 550
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.2937704918032787,
|
| 415 |
+
"grad_norm": 0.6028411388397217,
|
| 416 |
+
"learning_rate": 0.000148868028713418,
|
| 417 |
+
"loss": 0.6682120323181152,
|
| 418 |
+
"step": 560
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 0.29901639344262293,
|
| 422 |
+
"grad_norm": 0.6842379570007324,
|
| 423 |
+
"learning_rate": 0.0001477636664826063,
|
| 424 |
+
"loss": 0.6770430564880371,
|
| 425 |
+
"step": 570
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.3042622950819672,
|
| 429 |
+
"grad_norm": 1.2270914316177368,
|
| 430 |
+
"learning_rate": 0.0001466593042517946,
|
| 431 |
+
"loss": 0.704667329788208,
|
| 432 |
+
"step": 580
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.30950819672131147,
|
| 436 |
+
"grad_norm": 0.9267876148223877,
|
| 437 |
+
"learning_rate": 0.0001455549420209829,
|
| 438 |
+
"loss": 0.6503843784332275,
|
| 439 |
+
"step": 590
|
| 440 |
+
},
|
| 441 |
+
{
|
| 442 |
+
"epoch": 0.31475409836065577,
|
| 443 |
+
"grad_norm": 0.6021356582641602,
|
| 444 |
+
"learning_rate": 0.00014445057979017118,
|
| 445 |
+
"loss": 0.691249418258667,
|
| 446 |
+
"step": 600
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.32,
|
| 450 |
+
"grad_norm": 0.7987799048423767,
|
| 451 |
+
"learning_rate": 0.00014334621755935948,
|
| 452 |
+
"loss": 0.6753934383392334,
|
| 453 |
+
"step": 610
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.32524590163934425,
|
| 457 |
+
"grad_norm": 0.7932137250900269,
|
| 458 |
+
"learning_rate": 0.00014224185532854777,
|
| 459 |
+
"loss": 0.740220594406128,
|
| 460 |
+
"step": 620
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 0.33049180327868855,
|
| 464 |
+
"grad_norm": 0.6394649744033813,
|
| 465 |
+
"learning_rate": 0.00014113749309773606,
|
| 466 |
+
"loss": 0.6681424617767334,
|
| 467 |
+
"step": 630
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 0.3357377049180328,
|
| 471 |
+
"grad_norm": 0.6621270179748535,
|
| 472 |
+
"learning_rate": 0.00014003313086692436,
|
| 473 |
+
"loss": 0.6998953342437744,
|
| 474 |
+
"step": 640
|
| 475 |
+
},
|
| 476 |
+
{
|
| 477 |
+
"epoch": 0.34098360655737703,
|
| 478 |
+
"grad_norm": 0.6594683527946472,
|
| 479 |
+
"learning_rate": 0.00013892876863611265,
|
| 480 |
+
"loss": 0.6756977081298828,
|
| 481 |
+
"step": 650
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 0.3462295081967213,
|
| 485 |
+
"grad_norm": 0.6369149684906006,
|
| 486 |
+
"learning_rate": 0.00013782440640530095,
|
| 487 |
+
"loss": 0.6814823627471924,
|
| 488 |
+
"step": 660
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.35147540983606557,
|
| 492 |
+
"grad_norm": 0.5597897171974182,
|
| 493 |
+
"learning_rate": 0.00013672004417448924,
|
| 494 |
+
"loss": 0.6507920265197754,
|
| 495 |
+
"step": 670
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 0.35672131147540986,
|
| 499 |
+
"grad_norm": 0.6039972901344299,
|
| 500 |
+
"learning_rate": 0.00013561568194367753,
|
| 501 |
+
"loss": 0.6367014408111572,
|
| 502 |
+
"step": 680
|
| 503 |
+
},
|
| 504 |
+
{
|
| 505 |
+
"epoch": 0.3619672131147541,
|
| 506 |
+
"grad_norm": 0.5688148736953735,
|
| 507 |
+
"learning_rate": 0.00013451131971286583,
|
| 508 |
+
"loss": 0.6138595104217529,
|
| 509 |
+
"step": 690
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.36721311475409835,
|
| 513 |
+
"grad_norm": 0.6039051413536072,
|
| 514 |
+
"learning_rate": 0.00013340695748205412,
|
| 515 |
+
"loss": 0.6540034294128418,
|
| 516 |
+
"step": 700
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 0.37245901639344264,
|
| 520 |
+
"grad_norm": 0.6174374222755432,
|
| 521 |
+
"learning_rate": 0.00013230259525124242,
|
| 522 |
+
"loss": 0.6867434024810791,
|
| 523 |
+
"step": 710
|
| 524 |
+
},
|
| 525 |
+
{
|
| 526 |
+
"epoch": 0.3777049180327869,
|
| 527 |
+
"grad_norm": 0.8699930906295776,
|
| 528 |
+
"learning_rate": 0.0001311982330204307,
|
| 529 |
+
"loss": 0.6717788696289062,
|
| 530 |
+
"step": 720
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.3829508196721311,
|
| 534 |
+
"grad_norm": 0.7272820472717285,
|
| 535 |
+
"learning_rate": 0.000130093870789619,
|
| 536 |
+
"loss": 0.6900745868682862,
|
| 537 |
+
"step": 730
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.3881967213114754,
|
| 541 |
+
"grad_norm": 4.704572677612305,
|
| 542 |
+
"learning_rate": 0.0001289895085588073,
|
| 543 |
+
"loss": 0.6664741992950439,
|
| 544 |
+
"step": 740
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.39344262295081966,
|
| 548 |
+
"grad_norm": 0.7410299181938171,
|
| 549 |
+
"learning_rate": 0.0001278851463279956,
|
| 550 |
+
"loss": 0.6386402606964111,
|
| 551 |
+
"step": 750
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 0.39868852459016396,
|
| 555 |
+
"grad_norm": 0.6447997093200684,
|
| 556 |
+
"learning_rate": 0.00012678078409718389,
|
| 557 |
+
"loss": 0.656829833984375,
|
| 558 |
+
"step": 760
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.4039344262295082,
|
| 562 |
+
"grad_norm": 0.5744156837463379,
|
| 563 |
+
"learning_rate": 0.00012567642186637218,
|
| 564 |
+
"loss": 0.6654098510742188,
|
| 565 |
+
"step": 770
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 0.40918032786885244,
|
| 569 |
+
"grad_norm": 0.5341398119926453,
|
| 570 |
+
"learning_rate": 0.00012457205963556047,
|
| 571 |
+
"loss": 0.6406219005584717,
|
| 572 |
+
"step": 780
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.41442622950819674,
|
| 576 |
+
"grad_norm": 0.600540280342102,
|
| 577 |
+
"learning_rate": 0.00012346769740474877,
|
| 578 |
+
"loss": 0.659253454208374,
|
| 579 |
+
"step": 790
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.419672131147541,
|
| 583 |
+
"grad_norm": 0.7170221209526062,
|
| 584 |
+
"learning_rate": 0.00012236333517393706,
|
| 585 |
+
"loss": 0.6802017211914062,
|
| 586 |
+
"step": 800
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 0.4249180327868852,
|
| 590 |
+
"grad_norm": 0.7966761589050293,
|
| 591 |
+
"learning_rate": 0.00012125897294312535,
|
| 592 |
+
"loss": 0.6104480743408203,
|
| 593 |
+
"step": 810
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.4301639344262295,
|
| 597 |
+
"grad_norm": 0.671951413154602,
|
| 598 |
+
"learning_rate": 0.00012015461071231365,
|
| 599 |
+
"loss": 0.7058865070343018,
|
| 600 |
+
"step": 820
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.43540983606557376,
|
| 604 |
+
"grad_norm": 0.6206080913543701,
|
| 605 |
+
"learning_rate": 0.00011905024848150194,
|
| 606 |
+
"loss": 0.6379797458648682,
|
| 607 |
+
"step": 830
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 0.44065573770491806,
|
| 611 |
+
"grad_norm": 0.6439060568809509,
|
| 612 |
+
"learning_rate": 0.00011794588625069024,
|
| 613 |
+
"loss": 0.6559050559997559,
|
| 614 |
+
"step": 840
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.4459016393442623,
|
| 618 |
+
"grad_norm": 0.7650911808013916,
|
| 619 |
+
"learning_rate": 0.00011684152401987853,
|
| 620 |
+
"loss": 0.664661169052124,
|
| 621 |
+
"step": 850
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 0.45114754098360654,
|
| 625 |
+
"grad_norm": 0.6379468441009521,
|
| 626 |
+
"learning_rate": 0.00011573716178906681,
|
| 627 |
+
"loss": 0.6635408878326416,
|
| 628 |
+
"step": 860
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 0.45639344262295084,
|
| 632 |
+
"grad_norm": 0.6033166646957397,
|
| 633 |
+
"learning_rate": 0.00011463279955825512,
|
| 634 |
+
"loss": 0.6439225673675537,
|
| 635 |
+
"step": 870
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 0.4616393442622951,
|
| 639 |
+
"grad_norm": 0.6554288268089294,
|
| 640 |
+
"learning_rate": 0.0001135284373274434,
|
| 641 |
+
"loss": 0.6234912872314453,
|
| 642 |
+
"step": 880
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.4668852459016393,
|
| 646 |
+
"grad_norm": 0.7938706278800964,
|
| 647 |
+
"learning_rate": 0.0001124240750966317,
|
| 648 |
+
"loss": 0.6458295822143555,
|
| 649 |
+
"step": 890
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.4721311475409836,
|
| 653 |
+
"grad_norm": 0.7794424891471863,
|
| 654 |
+
"learning_rate": 0.00011131971286581999,
|
| 655 |
+
"loss": 0.6645576000213623,
|
| 656 |
+
"step": 900
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.47737704918032786,
|
| 660 |
+
"grad_norm": 0.6669194102287292,
|
| 661 |
+
"learning_rate": 0.0001102153506350083,
|
| 662 |
+
"loss": 0.706819725036621,
|
| 663 |
+
"step": 910
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 0.48262295081967216,
|
| 667 |
+
"grad_norm": 0.5930067896842957,
|
| 668 |
+
"learning_rate": 0.00010911098840419657,
|
| 669 |
+
"loss": 0.6377612113952636,
|
| 670 |
+
"step": 920
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 0.4878688524590164,
|
| 674 |
+
"grad_norm": 0.5960756540298462,
|
| 675 |
+
"learning_rate": 0.00010800662617338488,
|
| 676 |
+
"loss": 0.6585844993591309,
|
| 677 |
+
"step": 930
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 0.49311475409836064,
|
| 681 |
+
"grad_norm": 0.8185632228851318,
|
| 682 |
+
"learning_rate": 0.00010690226394257316,
|
| 683 |
+
"loss": 0.6509750366210938,
|
| 684 |
+
"step": 940
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.49836065573770494,
|
| 688 |
+
"grad_norm": 0.5078699588775635,
|
| 689 |
+
"learning_rate": 0.00010579790171176147,
|
| 690 |
+
"loss": 0.6630078315734863,
|
| 691 |
+
"step": 950
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.5036065573770492,
|
| 695 |
+
"grad_norm": 0.5847029685974121,
|
| 696 |
+
"learning_rate": 0.00010469353948094975,
|
| 697 |
+
"loss": 0.6561220169067383,
|
| 698 |
+
"step": 960
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.5088524590163934,
|
| 702 |
+
"grad_norm": 0.5837432146072388,
|
| 703 |
+
"learning_rate": 0.00010358917725013806,
|
| 704 |
+
"loss": 0.6355574607849122,
|
| 705 |
+
"step": 970
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.5140983606557377,
|
| 709 |
+
"grad_norm": 0.5932501554489136,
|
| 710 |
+
"learning_rate": 0.00010248481501932634,
|
| 711 |
+
"loss": 0.6037593364715577,
|
| 712 |
+
"step": 980
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 0.519344262295082,
|
| 716 |
+
"grad_norm": 0.6490769982337952,
|
| 717 |
+
"learning_rate": 0.00010138045278851464,
|
| 718 |
+
"loss": 0.7359431266784668,
|
| 719 |
+
"step": 990
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 0.5245901639344263,
|
| 723 |
+
"grad_norm": 0.8101118206977844,
|
| 724 |
+
"learning_rate": 0.00010027609055770292,
|
| 725 |
+
"loss": 0.5702367305755616,
|
| 726 |
+
"step": 1000
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 0.5245901639344263,
|
| 730 |
+
"eval_loss": 0.6413107514381409,
|
| 731 |
+
"eval_runtime": 327.3739,
|
| 732 |
+
"eval_samples_per_second": 34.459,
|
| 733 |
+
"eval_steps_per_second": 8.617,
|
| 734 |
+
"step": 1000
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.5245901639344263,
|
| 738 |
+
"step": 1000,
|
| 739 |
+
"test_best_wer": 1.068743341281518,
|
| 740 |
+
"test_exact_match": 0.0,
|
| 741 |
+
"test_hallucination_rate": 0.2734375,
|
| 742 |
+
"test_wer": 1.068743341281518
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 0.5298360655737705,
|
| 746 |
+
"grad_norm": 0.48935407400131226,
|
| 747 |
+
"learning_rate": 9.917172832689122e-05,
|
| 748 |
+
"loss": 0.7065565586090088,
|
| 749 |
+
"step": 1010
|
| 750 |
+
},
|
| 751 |
+
{
|
| 752 |
+
"epoch": 0.5350819672131147,
|
| 753 |
+
"grad_norm": 0.5701637864112854,
|
| 754 |
+
"learning_rate": 9.806736609607951e-05,
|
| 755 |
+
"loss": 0.632069206237793,
|
| 756 |
+
"step": 1020
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 0.540327868852459,
|
| 760 |
+
"grad_norm": 0.6583526134490967,
|
| 761 |
+
"learning_rate": 9.69630038652678e-05,
|
| 762 |
+
"loss": 0.6403782367706299,
|
| 763 |
+
"step": 1030
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 0.5455737704918033,
|
| 767 |
+
"grad_norm": 0.6018849611282349,
|
| 768 |
+
"learning_rate": 9.58586416344561e-05,
|
| 769 |
+
"loss": 0.6802470207214355,
|
| 770 |
+
"step": 1040
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.5508196721311476,
|
| 774 |
+
"grad_norm": 1.0071996450424194,
|
| 775 |
+
"learning_rate": 9.47542794036444e-05,
|
| 776 |
+
"loss": 0.644084882736206,
|
| 777 |
+
"step": 1050
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.5560655737704918,
|
| 781 |
+
"grad_norm": 0.7839695811271667,
|
| 782 |
+
"learning_rate": 9.364991717283269e-05,
|
| 783 |
+
"loss": 0.6707189083099365,
|
| 784 |
+
"step": 1060
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 0.5613114754098361,
|
| 788 |
+
"grad_norm": 0.6655315160751343,
|
| 789 |
+
"learning_rate": 9.254555494202098e-05,
|
| 790 |
+
"loss": 0.6463171958923339,
|
| 791 |
+
"step": 1070
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 0.5665573770491803,
|
| 795 |
+
"grad_norm": 0.6828731298446655,
|
| 796 |
+
"learning_rate": 9.144119271120928e-05,
|
| 797 |
+
"loss": 0.5902272701263428,
|
| 798 |
+
"step": 1080
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 0.5718032786885245,
|
| 802 |
+
"grad_norm": 0.5706672072410583,
|
| 803 |
+
"learning_rate": 9.033683048039757e-05,
|
| 804 |
+
"loss": 0.6675863265991211,
|
| 805 |
+
"step": 1090
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 0.5770491803278689,
|
| 809 |
+
"grad_norm": 0.5320814251899719,
|
| 810 |
+
"learning_rate": 8.923246824958586e-05,
|
| 811 |
+
"loss": 0.6380939483642578,
|
| 812 |
+
"step": 1100
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.5822950819672131,
|
| 816 |
+
"grad_norm": 2.452364921569824,
|
| 817 |
+
"learning_rate": 8.812810601877416e-05,
|
| 818 |
+
"loss": 0.6228958129882812,
|
| 819 |
+
"step": 1110
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.5875409836065574,
|
| 823 |
+
"grad_norm": 0.6400389671325684,
|
| 824 |
+
"learning_rate": 8.702374378796245e-05,
|
| 825 |
+
"loss": 0.6083290576934814,
|
| 826 |
+
"step": 1120
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.5927868852459016,
|
| 830 |
+
"grad_norm": 0.562908947467804,
|
| 831 |
+
"learning_rate": 8.591938155715074e-05,
|
| 832 |
+
"loss": 0.6654650688171386,
|
| 833 |
+
"step": 1130
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.5980327868852459,
|
| 837 |
+
"grad_norm": 0.9055919051170349,
|
| 838 |
+
"learning_rate": 8.481501932633904e-05,
|
| 839 |
+
"loss": 0.6651034832000733,
|
| 840 |
+
"step": 1140
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 0.6032786885245902,
|
| 844 |
+
"grad_norm": 0.5761740803718567,
|
| 845 |
+
"learning_rate": 8.371065709552733e-05,
|
| 846 |
+
"loss": 0.6683104991912842,
|
| 847 |
+
"step": 1150
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 0.6085245901639345,
|
| 851 |
+
"grad_norm": 0.6321728825569153,
|
| 852 |
+
"learning_rate": 8.260629486471563e-05,
|
| 853 |
+
"loss": 0.6309406280517578,
|
| 854 |
+
"step": 1160
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.6137704918032787,
|
| 858 |
+
"grad_norm": 0.7833631038665771,
|
| 859 |
+
"learning_rate": 8.150193263390392e-05,
|
| 860 |
+
"loss": 0.5802960872650147,
|
| 861 |
+
"step": 1170
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.6190163934426229,
|
| 865 |
+
"grad_norm": 0.618356466293335,
|
| 866 |
+
"learning_rate": 8.039757040309221e-05,
|
| 867 |
+
"loss": 0.648448371887207,
|
| 868 |
+
"step": 1180
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"epoch": 0.6242622950819672,
|
| 872 |
+
"grad_norm": 0.5821840167045593,
|
| 873 |
+
"learning_rate": 7.929320817228051e-05,
|
| 874 |
+
"loss": 0.6359022617340088,
|
| 875 |
+
"step": 1190
|
| 876 |
+
},
|
| 877 |
+
{
|
| 878 |
+
"epoch": 0.6295081967213115,
|
| 879 |
+
"grad_norm": 0.5751839280128479,
|
| 880 |
+
"learning_rate": 7.81888459414688e-05,
|
| 881 |
+
"loss": 0.6848149776458741,
|
| 882 |
+
"step": 1200
|
| 883 |
+
},
|
| 884 |
+
{
|
| 885 |
+
"epoch": 0.6347540983606558,
|
| 886 |
+
"grad_norm": 0.6764397025108337,
|
| 887 |
+
"learning_rate": 7.70844837106571e-05,
|
| 888 |
+
"loss": 0.6194316864013671,
|
| 889 |
+
"step": 1210
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.64,
|
| 893 |
+
"grad_norm": 0.7781165242195129,
|
| 894 |
+
"learning_rate": 7.598012147984539e-05,
|
| 895 |
+
"loss": 0.5850466728210449,
|
| 896 |
+
"step": 1220
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.6452459016393443,
|
| 900 |
+
"grad_norm": 0.5429039597511292,
|
| 901 |
+
"learning_rate": 7.487575924903368e-05,
|
| 902 |
+
"loss": 0.6222261905670166,
|
| 903 |
+
"step": 1230
|
| 904 |
+
},
|
| 905 |
+
{
|
| 906 |
+
"epoch": 0.6504918032786885,
|
| 907 |
+
"grad_norm": 0.592396080493927,
|
| 908 |
+
"learning_rate": 7.377139701822198e-05,
|
| 909 |
+
"loss": 0.5790858268737793,
|
| 910 |
+
"step": 1240
|
| 911 |
+
},
|
| 912 |
+
{
|
| 913 |
+
"epoch": 0.6557377049180327,
|
| 914 |
+
"grad_norm": 0.6150607466697693,
|
| 915 |
+
"learning_rate": 7.266703478741027e-05,
|
| 916 |
+
"loss": 0.5980223178863525,
|
| 917 |
+
"step": 1250
|
| 918 |
+
},
|
| 919 |
+
{
|
| 920 |
+
"epoch": 0.6609836065573771,
|
| 921 |
+
"grad_norm": 0.6551251411437988,
|
| 922 |
+
"learning_rate": 7.156267255659857e-05,
|
| 923 |
+
"loss": 0.5977930068969727,
|
| 924 |
+
"step": 1260
|
| 925 |
+
},
|
| 926 |
+
{
|
| 927 |
+
"epoch": 0.6662295081967213,
|
| 928 |
+
"grad_norm": 0.6054208874702454,
|
| 929 |
+
"learning_rate": 7.045831032578686e-05,
|
| 930 |
+
"loss": 0.5855415821075439,
|
| 931 |
+
"step": 1270
|
| 932 |
+
},
|
| 933 |
+
{
|
| 934 |
+
"epoch": 0.6714754098360656,
|
| 935 |
+
"grad_norm": 0.502633810043335,
|
| 936 |
+
"learning_rate": 6.935394809497515e-05,
|
| 937 |
+
"loss": 0.5847459316253663,
|
| 938 |
+
"step": 1280
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.6767213114754098,
|
| 942 |
+
"grad_norm": 0.6111732125282288,
|
| 943 |
+
"learning_rate": 6.824958586416345e-05,
|
| 944 |
+
"loss": 0.6229658603668213,
|
| 945 |
+
"step": 1290
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.6819672131147541,
|
| 949 |
+
"grad_norm": 0.4827969968318939,
|
| 950 |
+
"learning_rate": 6.714522363335174e-05,
|
| 951 |
+
"loss": 0.6103329181671142,
|
| 952 |
+
"step": 1300
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 0.6872131147540984,
|
| 956 |
+
"grad_norm": 0.780549943447113,
|
| 957 |
+
"learning_rate": 6.604086140254003e-05,
|
| 958 |
+
"loss": 0.6113051891326904,
|
| 959 |
+
"step": 1310
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 0.6924590163934427,
|
| 963 |
+
"grad_norm": 0.7107716798782349,
|
| 964 |
+
"learning_rate": 6.493649917172833e-05,
|
| 965 |
+
"loss": 0.6017592906951904,
|
| 966 |
+
"step": 1320
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 0.6977049180327869,
|
| 970 |
+
"grad_norm": 0.6018385291099548,
|
| 971 |
+
"learning_rate": 6.383213694091662e-05,
|
| 972 |
+
"loss": 0.648677921295166,
|
| 973 |
+
"step": 1330
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 0.7029508196721311,
|
| 977 |
+
"grad_norm": 0.6865012049674988,
|
| 978 |
+
"learning_rate": 6.272777471010492e-05,
|
| 979 |
+
"loss": 0.5954135417938232,
|
| 980 |
+
"step": 1340
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.7081967213114754,
|
| 984 |
+
"grad_norm": 0.6098220348358154,
|
| 985 |
+
"learning_rate": 6.162341247929321e-05,
|
| 986 |
+
"loss": 0.6382476329803467,
|
| 987 |
+
"step": 1350
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.7134426229508197,
|
| 991 |
+
"grad_norm": 0.6140968799591064,
|
| 992 |
+
"learning_rate": 6.0519050248481504e-05,
|
| 993 |
+
"loss": 0.5865953445434571,
|
| 994 |
+
"step": 1360
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 0.718688524590164,
|
| 998 |
+
"grad_norm": 0.5503070950508118,
|
| 999 |
+
"learning_rate": 5.94146880176698e-05,
|
| 1000 |
+
"loss": 0.5596487998962403,
|
| 1001 |
+
"step": 1370
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.7239344262295082,
|
| 1005 |
+
"grad_norm": 0.6459322571754456,
|
| 1006 |
+
"learning_rate": 5.831032578685809e-05,
|
| 1007 |
+
"loss": 0.577778673171997,
|
| 1008 |
+
"step": 1380
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 0.7291803278688525,
|
| 1012 |
+
"grad_norm": 0.5784876942634583,
|
| 1013 |
+
"learning_rate": 5.7205963556046386e-05,
|
| 1014 |
+
"loss": 0.6004335403442382,
|
| 1015 |
+
"step": 1390
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 0.7344262295081967,
|
| 1019 |
+
"grad_norm": 0.5703299045562744,
|
| 1020 |
+
"learning_rate": 5.610160132523468e-05,
|
| 1021 |
+
"loss": 0.6232605457305909,
|
| 1022 |
+
"step": 1400
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 0.7396721311475409,
|
| 1026 |
+
"grad_norm": 0.9156650900840759,
|
| 1027 |
+
"learning_rate": 5.4997239094422967e-05,
|
| 1028 |
+
"loss": 0.6033765792846679,
|
| 1029 |
+
"step": 1410
|
| 1030 |
+
},
|
| 1031 |
+
{
|
| 1032 |
+
"epoch": 0.7449180327868853,
|
| 1033 |
+
"grad_norm": 0.6295714378356934,
|
| 1034 |
+
"learning_rate": 5.389287686361126e-05,
|
| 1035 |
+
"loss": 0.6655001163482666,
|
| 1036 |
+
"step": 1420
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 0.7501639344262295,
|
| 1040 |
+
"grad_norm": 0.6699951887130737,
|
| 1041 |
+
"learning_rate": 5.2788514632799554e-05,
|
| 1042 |
+
"loss": 0.6214305877685546,
|
| 1043 |
+
"step": 1430
|
| 1044 |
+
},
|
| 1045 |
+
{
|
| 1046 |
+
"epoch": 0.7554098360655738,
|
| 1047 |
+
"grad_norm": 0.6853722333908081,
|
| 1048 |
+
"learning_rate": 5.168415240198785e-05,
|
| 1049 |
+
"loss": 0.5730300903320312,
|
| 1050 |
+
"step": 1440
|
| 1051 |
+
},
|
| 1052 |
+
{
|
| 1053 |
+
"epoch": 0.760655737704918,
|
| 1054 |
+
"grad_norm": 0.6951788067817688,
|
| 1055 |
+
"learning_rate": 5.057979017117614e-05,
|
| 1056 |
+
"loss": 0.6502082824707032,
|
| 1057 |
+
"step": 1450
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 0.7659016393442623,
|
| 1061 |
+
"grad_norm": 0.6293227672576904,
|
| 1062 |
+
"learning_rate": 4.947542794036444e-05,
|
| 1063 |
+
"loss": 0.5680208683013916,
|
| 1064 |
+
"step": 1460
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.7711475409836066,
|
| 1068 |
+
"grad_norm": 0.5696614384651184,
|
| 1069 |
+
"learning_rate": 4.8371065709552737e-05,
|
| 1070 |
+
"loss": 0.5762951850891114,
|
| 1071 |
+
"step": 1470
|
| 1072 |
+
},
|
| 1073 |
+
{
|
| 1074 |
+
"epoch": 0.7763934426229508,
|
| 1075 |
+
"grad_norm": 0.7280446290969849,
|
| 1076 |
+
"learning_rate": 4.726670347874103e-05,
|
| 1077 |
+
"loss": 0.6193453788757324,
|
| 1078 |
+
"step": 1480
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 0.7816393442622951,
|
| 1082 |
+
"grad_norm": 0.6729447245597839,
|
| 1083 |
+
"learning_rate": 4.6162341247929324e-05,
|
| 1084 |
+
"loss": 0.6109768867492675,
|
| 1085 |
+
"step": 1490
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"epoch": 0.7868852459016393,
|
| 1089 |
+
"grad_norm": 0.5846381187438965,
|
| 1090 |
+
"learning_rate": 4.505797901711762e-05,
|
| 1091 |
+
"loss": 0.6103529930114746,
|
| 1092 |
+
"step": 1500
|
| 1093 |
+
},
|
| 1094 |
+
{
|
| 1095 |
+
"epoch": 0.7868852459016393,
|
| 1096 |
+
"eval_loss": 0.6120977997779846,
|
| 1097 |
+
"eval_runtime": 328.0023,
|
| 1098 |
+
"eval_samples_per_second": 34.393,
|
| 1099 |
+
"eval_steps_per_second": 8.601,
|
| 1100 |
+
"step": 1500
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.7868852459016393,
|
| 1104 |
+
"step": 1500,
|
| 1105 |
+
"test_best_wer": 0.6988483587844351,
|
| 1106 |
+
"test_exact_match": 0.0,
|
| 1107 |
+
"test_hallucination_rate": 0.1484375,
|
| 1108 |
+
"test_wer": 0.6988483587844351
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.7921311475409836,
|
| 1112 |
+
"grad_norm": 0.6908854842185974,
|
| 1113 |
+
"learning_rate": 4.395361678630591e-05,
|
| 1114 |
+
"loss": 0.5708508491516113,
|
| 1115 |
+
"step": 1510
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.7973770491803279,
|
| 1119 |
+
"grad_norm": 0.5843254923820496,
|
| 1120 |
+
"learning_rate": 4.2849254555494206e-05,
|
| 1121 |
+
"loss": 0.5473520755767822,
|
| 1122 |
+
"step": 1520
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.8026229508196722,
|
| 1126 |
+
"grad_norm": 0.5113570690155029,
|
| 1127 |
+
"learning_rate": 4.17448923246825e-05,
|
| 1128 |
+
"loss": 0.644585657119751,
|
| 1129 |
+
"step": 1530
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.8078688524590164,
|
| 1133 |
+
"grad_norm": 0.704180121421814,
|
| 1134 |
+
"learning_rate": 4.0640530093870794e-05,
|
| 1135 |
+
"loss": 0.609114122390747,
|
| 1136 |
+
"step": 1540
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.8131147540983606,
|
| 1140 |
+
"grad_norm": 0.6005797982215881,
|
| 1141 |
+
"learning_rate": 3.953616786305909e-05,
|
| 1142 |
+
"loss": 0.6127142429351806,
|
| 1143 |
+
"step": 1550
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.8183606557377049,
|
| 1147 |
+
"grad_norm": 0.6694676280021667,
|
| 1148 |
+
"learning_rate": 3.843180563224738e-05,
|
| 1149 |
+
"loss": 0.5431616306304932,
|
| 1150 |
+
"step": 1560
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.8236065573770491,
|
| 1154 |
+
"grad_norm": 4.152336120605469,
|
| 1155 |
+
"learning_rate": 3.7327443401435675e-05,
|
| 1156 |
+
"loss": 0.6213226318359375,
|
| 1157 |
+
"step": 1570
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.8288524590163935,
|
| 1161 |
+
"grad_norm": 0.7689782381057739,
|
| 1162 |
+
"learning_rate": 3.622308117062397e-05,
|
| 1163 |
+
"loss": 0.6467239379882812,
|
| 1164 |
+
"step": 1580
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.8340983606557377,
|
| 1168 |
+
"grad_norm": 1.6682908535003662,
|
| 1169 |
+
"learning_rate": 3.511871893981226e-05,
|
| 1170 |
+
"loss": 0.6183376789093018,
|
| 1171 |
+
"step": 1590
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.839344262295082,
|
| 1175 |
+
"grad_norm": 0.5368894934654236,
|
| 1176 |
+
"learning_rate": 3.401435670900056e-05,
|
| 1177 |
+
"loss": 0.5576130390167237,
|
| 1178 |
+
"step": 1600
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.8445901639344262,
|
| 1182 |
+
"grad_norm": 0.7344756126403809,
|
| 1183 |
+
"learning_rate": 3.290999447818885e-05,
|
| 1184 |
+
"loss": 0.6612845897674561,
|
| 1185 |
+
"step": 1610
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.8498360655737704,
|
| 1189 |
+
"grad_norm": 0.7366407513618469,
|
| 1190 |
+
"learning_rate": 3.1805632247377145e-05,
|
| 1191 |
+
"loss": 0.5778892993927002,
|
| 1192 |
+
"step": 1620
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.8550819672131148,
|
| 1196 |
+
"grad_norm": 0.7209796905517578,
|
| 1197 |
+
"learning_rate": 3.070127001656544e-05,
|
| 1198 |
+
"loss": 0.6044011592864991,
|
| 1199 |
+
"step": 1630
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 0.860327868852459,
|
| 1203 |
+
"grad_norm": 0.744916558265686,
|
| 1204 |
+
"learning_rate": 2.959690778575373e-05,
|
| 1205 |
+
"loss": 0.6034855842590332,
|
| 1206 |
+
"step": 1640
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 0.8655737704918033,
|
| 1210 |
+
"grad_norm": 0.5565060377120972,
|
| 1211 |
+
"learning_rate": 2.8492545554942023e-05,
|
| 1212 |
+
"loss": 0.6295875549316406,
|
| 1213 |
+
"step": 1650
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.8708196721311475,
|
| 1217 |
+
"grad_norm": 0.5516486167907715,
|
| 1218 |
+
"learning_rate": 2.7388183324130317e-05,
|
| 1219 |
+
"loss": 0.6283913612365722,
|
| 1220 |
+
"step": 1660
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.8760655737704918,
|
| 1224 |
+
"grad_norm": 0.6224448084831238,
|
| 1225 |
+
"learning_rate": 2.628382109331861e-05,
|
| 1226 |
+
"loss": 0.5944302082061768,
|
| 1227 |
+
"step": 1670
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 0.8813114754098361,
|
| 1231 |
+
"grad_norm": 0.5552781224250793,
|
| 1232 |
+
"learning_rate": 2.5179458862506904e-05,
|
| 1233 |
+
"loss": 0.5666909217834473,
|
| 1234 |
+
"step": 1680
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.8865573770491804,
|
| 1238 |
+
"grad_norm": 0.5582685470581055,
|
| 1239 |
+
"learning_rate": 2.4075096631695195e-05,
|
| 1240 |
+
"loss": 0.6030192375183105,
|
| 1241 |
+
"step": 1690
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.8918032786885246,
|
| 1245 |
+
"grad_norm": 0.7704609036445618,
|
| 1246 |
+
"learning_rate": 2.297073440088349e-05,
|
| 1247 |
+
"loss": 0.6008219242095947,
|
| 1248 |
+
"step": 1700
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.8970491803278688,
|
| 1252 |
+
"grad_norm": 0.6567322015762329,
|
| 1253 |
+
"learning_rate": 2.1866372170071783e-05,
|
| 1254 |
+
"loss": 0.6032386302947998,
|
| 1255 |
+
"step": 1710
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 0.9022950819672131,
|
| 1259 |
+
"grad_norm": 0.537177562713623,
|
| 1260 |
+
"learning_rate": 2.0762009939260077e-05,
|
| 1261 |
+
"loss": 0.5888541221618653,
|
| 1262 |
+
"step": 1720
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.9075409836065573,
|
| 1266 |
+
"grad_norm": 0.7722596526145935,
|
| 1267 |
+
"learning_rate": 1.965764770844837e-05,
|
| 1268 |
+
"loss": 0.6460831642150879,
|
| 1269 |
+
"step": 1730
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.9127868852459017,
|
| 1273 |
+
"grad_norm": 1.7876378297805786,
|
| 1274 |
+
"learning_rate": 1.8553285477636664e-05,
|
| 1275 |
+
"loss": 0.5897231101989746,
|
| 1276 |
+
"step": 1740
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 0.9180327868852459,
|
| 1280 |
+
"grad_norm": 0.5136721730232239,
|
| 1281 |
+
"learning_rate": 1.7448923246824958e-05,
|
| 1282 |
+
"loss": 0.5294720649719238,
|
| 1283 |
+
"step": 1750
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 0.9232786885245902,
|
| 1287 |
+
"grad_norm": 0.6075170040130615,
|
| 1288 |
+
"learning_rate": 1.6344561016013255e-05,
|
| 1289 |
+
"loss": 0.5763162612915039,
|
| 1290 |
+
"step": 1760
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.9285245901639344,
|
| 1294 |
+
"grad_norm": 0.5250259637832642,
|
| 1295 |
+
"learning_rate": 1.5240198785201548e-05,
|
| 1296 |
+
"loss": 0.5351035118103027,
|
| 1297 |
+
"step": 1770
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 0.9337704918032786,
|
| 1301 |
+
"grad_norm": 0.7039772868156433,
|
| 1302 |
+
"learning_rate": 1.4135836554389841e-05,
|
| 1303 |
+
"loss": 0.6187419414520263,
|
| 1304 |
+
"step": 1780
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.939016393442623,
|
| 1308 |
+
"grad_norm": 0.5285559296607971,
|
| 1309 |
+
"learning_rate": 1.3031474323578135e-05,
|
| 1310 |
+
"loss": 0.6073878288269043,
|
| 1311 |
+
"step": 1790
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 0.9442622950819672,
|
| 1315 |
+
"grad_norm": 0.7015911936759949,
|
| 1316 |
+
"learning_rate": 1.1927112092766428e-05,
|
| 1317 |
+
"loss": 0.6200543403625488,
|
| 1318 |
+
"step": 1800
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 0.9495081967213115,
|
| 1322 |
+
"grad_norm": 0.5723676085472107,
|
| 1323 |
+
"learning_rate": 1.0822749861954721e-05,
|
| 1324 |
+
"loss": 0.5787985324859619,
|
| 1325 |
+
"step": 1810
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 0.9547540983606557,
|
| 1329 |
+
"grad_norm": 0.6535468697547913,
|
| 1330 |
+
"learning_rate": 9.718387631143015e-06,
|
| 1331 |
+
"loss": 0.6332112789154053,
|
| 1332 |
+
"step": 1820
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.96,
|
| 1336 |
+
"grad_norm": 0.6413028836250305,
|
| 1337 |
+
"learning_rate": 8.61402540033131e-06,
|
| 1338 |
+
"loss": 0.57379150390625,
|
| 1339 |
+
"step": 1830
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 0.9652459016393443,
|
| 1343 |
+
"grad_norm": 0.8077524304389954,
|
| 1344 |
+
"learning_rate": 7.509663169519603e-06,
|
| 1345 |
+
"loss": 0.5872003555297851,
|
| 1346 |
+
"step": 1840
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.9704918032786886,
|
| 1350 |
+
"grad_norm": 0.5533561706542969,
|
| 1351 |
+
"learning_rate": 6.405300938707896e-06,
|
| 1352 |
+
"loss": 0.5622630596160889,
|
| 1353 |
+
"step": 1850
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.9757377049180328,
|
| 1357 |
+
"grad_norm": 0.6691303253173828,
|
| 1358 |
+
"learning_rate": 5.30093870789619e-06,
|
| 1359 |
+
"loss": 0.5612599372863769,
|
| 1360 |
+
"step": 1860
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.980983606557377,
|
| 1364 |
+
"grad_norm": 0.7582765221595764,
|
| 1365 |
+
"learning_rate": 4.196576477084484e-06,
|
| 1366 |
+
"loss": 0.5620477676391602,
|
| 1367 |
+
"step": 1870
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 0.9862295081967213,
|
| 1371 |
+
"grad_norm": 0.6176788806915283,
|
| 1372 |
+
"learning_rate": 3.0922142462727777e-06,
|
| 1373 |
+
"loss": 0.59493989944458,
|
| 1374 |
+
"step": 1880
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.9914754098360655,
|
| 1378 |
+
"grad_norm": 0.6041266918182373,
|
| 1379 |
+
"learning_rate": 1.9878520154610715e-06,
|
| 1380 |
+
"loss": 0.5545070171356201,
|
| 1381 |
+
"step": 1890
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 0.9967213114754099,
|
| 1385 |
+
"grad_norm": 0.5746496319770813,
|
| 1386 |
+
"learning_rate": 8.834897846493651e-07,
|
| 1387 |
+
"loss": 0.6181450843811035,
|
| 1388 |
+
"step": 1900
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.0,
|
| 1392 |
+
"eval_loss": 0.5994250774383545,
|
| 1393 |
+
"eval_runtime": 326.9291,
|
| 1394 |
+
"eval_samples_per_second": 34.506,
|
| 1395 |
+
"eval_steps_per_second": 8.629,
|
| 1396 |
+
"step": 1907
|
| 1397 |
+
}
|
| 1398 |
+
],
|
| 1399 |
+
"logging_steps": 10,
|
| 1400 |
+
"max_steps": 1907,
|
| 1401 |
+
"num_input_tokens_seen": 0,
|
| 1402 |
+
"num_train_epochs": 1,
|
| 1403 |
+
"save_steps": 500,
|
| 1404 |
+
"stateful_callbacks": {
|
| 1405 |
+
"TrainerControl": {
|
| 1406 |
+
"args": {
|
| 1407 |
+
"should_epoch_stop": false,
|
| 1408 |
+
"should_evaluate": false,
|
| 1409 |
+
"should_log": false,
|
| 1410 |
+
"should_save": true,
|
| 1411 |
+
"should_training_stop": true
|
| 1412 |
+
},
|
| 1413 |
+
"attributes": {}
|
| 1414 |
+
}
|
| 1415 |
+
},
|
| 1416 |
+
"total_flos": 1.095908244879506e+17,
|
| 1417 |
+
"train_batch_size": 2,
|
| 1418 |
+
"trial_name": null,
|
| 1419 |
+
"trial_params": null
|
| 1420 |
+
}
|
gemma-3-1b-it/{checkpoint-500 → checkpoint-1907}/training_args.bin
RENAMED
|
File without changes
|
gemma-3-1b-it/checkpoint-500/trainer_state.json
DELETED
|
@@ -1,392 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": null,
|
| 3 |
-
"best_metric": null,
|
| 4 |
-
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.26229508196721313,
|
| 6 |
-
"eval_steps": 500,
|
| 7 |
-
"global_step": 500,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.005245901639344262,
|
| 14 |
-
"grad_norm": 5.850006580352783,
|
| 15 |
-
"learning_rate": 1.8750000000000002e-05,
|
| 16 |
-
"loss": 1.626298713684082,
|
| 17 |
-
"step": 10
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.010491803278688525,
|
| 21 |
-
"grad_norm": 1.7961169481277466,
|
| 22 |
-
"learning_rate": 3.958333333333333e-05,
|
| 23 |
-
"loss": 1.3221087455749512,
|
| 24 |
-
"step": 20
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.015737704918032787,
|
| 28 |
-
"grad_norm": 1.4881651401519775,
|
| 29 |
-
"learning_rate": 6.041666666666667e-05,
|
| 30 |
-
"loss": 1.1436286926269532,
|
| 31 |
-
"step": 30
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.02098360655737705,
|
| 35 |
-
"grad_norm": 1.0592900514602661,
|
| 36 |
-
"learning_rate": 8.125000000000001e-05,
|
| 37 |
-
"loss": 0.9118364334106446,
|
| 38 |
-
"step": 40
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.02622950819672131,
|
| 42 |
-
"grad_norm": 1.3136152029037476,
|
| 43 |
-
"learning_rate": 0.00010208333333333333,
|
| 44 |
-
"loss": 0.9902483940124511,
|
| 45 |
-
"step": 50
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.031475409836065574,
|
| 49 |
-
"grad_norm": 1.0517693758010864,
|
| 50 |
-
"learning_rate": 0.00012291666666666668,
|
| 51 |
-
"loss": 0.9533065795898438,
|
| 52 |
-
"step": 60
|
| 53 |
-
},
|
| 54 |
-
{
|
| 55 |
-
"epoch": 0.036721311475409836,
|
| 56 |
-
"grad_norm": 1.0763620138168335,
|
| 57 |
-
"learning_rate": 0.00014375,
|
| 58 |
-
"loss": 0.9590832710266113,
|
| 59 |
-
"step": 70
|
| 60 |
-
},
|
| 61 |
-
{
|
| 62 |
-
"epoch": 0.0419672131147541,
|
| 63 |
-
"grad_norm": 1.126420497894287,
|
| 64 |
-
"learning_rate": 0.00016458333333333334,
|
| 65 |
-
"loss": 0.8696105003356933,
|
| 66 |
-
"step": 80
|
| 67 |
-
},
|
| 68 |
-
{
|
| 69 |
-
"epoch": 0.04721311475409836,
|
| 70 |
-
"grad_norm": 1.0479974746704102,
|
| 71 |
-
"learning_rate": 0.00018541666666666668,
|
| 72 |
-
"loss": 0.8585530281066894,
|
| 73 |
-
"step": 90
|
| 74 |
-
},
|
| 75 |
-
{
|
| 76 |
-
"epoch": 0.05245901639344262,
|
| 77 |
-
"grad_norm": 0.9453676342964172,
|
| 78 |
-
"learning_rate": 0.0001996686913307565,
|
| 79 |
-
"loss": 0.81344633102417,
|
| 80 |
-
"step": 100
|
| 81 |
-
},
|
| 82 |
-
{
|
| 83 |
-
"epoch": 0.057704918032786885,
|
| 84 |
-
"grad_norm": 0.9837343692779541,
|
| 85 |
-
"learning_rate": 0.0001985643290999448,
|
| 86 |
-
"loss": 0.9181200981140136,
|
| 87 |
-
"step": 110
|
| 88 |
-
},
|
| 89 |
-
{
|
| 90 |
-
"epoch": 0.06295081967213115,
|
| 91 |
-
"grad_norm": 0.8958956003189087,
|
| 92 |
-
"learning_rate": 0.00019745996686913308,
|
| 93 |
-
"loss": 0.826357650756836,
|
| 94 |
-
"step": 120
|
| 95 |
-
},
|
| 96 |
-
{
|
| 97 |
-
"epoch": 0.06819672131147542,
|
| 98 |
-
"grad_norm": 0.9927870631217957,
|
| 99 |
-
"learning_rate": 0.0001963556046383214,
|
| 100 |
-
"loss": 0.8804688453674316,
|
| 101 |
-
"step": 130
|
| 102 |
-
},
|
| 103 |
-
{
|
| 104 |
-
"epoch": 0.07344262295081967,
|
| 105 |
-
"grad_norm": 0.8682142496109009,
|
| 106 |
-
"learning_rate": 0.00019525124240750967,
|
| 107 |
-
"loss": 0.8082976341247559,
|
| 108 |
-
"step": 140
|
| 109 |
-
},
|
| 110 |
-
{
|
| 111 |
-
"epoch": 0.07868852459016394,
|
| 112 |
-
"grad_norm": 0.837242841720581,
|
| 113 |
-
"learning_rate": 0.00019414688017669799,
|
| 114 |
-
"loss": 0.8015812873840332,
|
| 115 |
-
"step": 150
|
| 116 |
-
},
|
| 117 |
-
{
|
| 118 |
-
"epoch": 0.0839344262295082,
|
| 119 |
-
"grad_norm": 1.099856972694397,
|
| 120 |
-
"learning_rate": 0.00019304251794588625,
|
| 121 |
-
"loss": 0.7791732311248779,
|
| 122 |
-
"step": 160
|
| 123 |
-
},
|
| 124 |
-
{
|
| 125 |
-
"epoch": 0.08918032786885247,
|
| 126 |
-
"grad_norm": 0.9170840978622437,
|
| 127 |
-
"learning_rate": 0.00019193815571507457,
|
| 128 |
-
"loss": 0.8227507591247558,
|
| 129 |
-
"step": 170
|
| 130 |
-
},
|
| 131 |
-
{
|
| 132 |
-
"epoch": 0.09442622950819672,
|
| 133 |
-
"grad_norm": 0.8296833038330078,
|
| 134 |
-
"learning_rate": 0.00019083379348426284,
|
| 135 |
-
"loss": 0.7757059097290039,
|
| 136 |
-
"step": 180
|
| 137 |
-
},
|
| 138 |
-
{
|
| 139 |
-
"epoch": 0.09967213114754099,
|
| 140 |
-
"grad_norm": 0.7342070937156677,
|
| 141 |
-
"learning_rate": 0.00018972943125345116,
|
| 142 |
-
"loss": 0.7799826622009277,
|
| 143 |
-
"step": 190
|
| 144 |
-
},
|
| 145 |
-
{
|
| 146 |
-
"epoch": 0.10491803278688525,
|
| 147 |
-
"grad_norm": 0.7804243564605713,
|
| 148 |
-
"learning_rate": 0.00018862506902263943,
|
| 149 |
-
"loss": 0.7913103580474854,
|
| 150 |
-
"step": 200
|
| 151 |
-
},
|
| 152 |
-
{
|
| 153 |
-
"epoch": 0.11016393442622952,
|
| 154 |
-
"grad_norm": 0.7638176083564758,
|
| 155 |
-
"learning_rate": 0.00018752070679182775,
|
| 156 |
-
"loss": 0.7876776695251465,
|
| 157 |
-
"step": 210
|
| 158 |
-
},
|
| 159 |
-
{
|
| 160 |
-
"epoch": 0.11540983606557377,
|
| 161 |
-
"grad_norm": 0.6659265160560608,
|
| 162 |
-
"learning_rate": 0.00018641634456101602,
|
| 163 |
-
"loss": 0.7114231586456299,
|
| 164 |
-
"step": 220
|
| 165 |
-
},
|
| 166 |
-
{
|
| 167 |
-
"epoch": 0.12065573770491804,
|
| 168 |
-
"grad_norm": 0.6473881602287292,
|
| 169 |
-
"learning_rate": 0.00018531198233020434,
|
| 170 |
-
"loss": 0.7239602088928223,
|
| 171 |
-
"step": 230
|
| 172 |
-
},
|
| 173 |
-
{
|
| 174 |
-
"epoch": 0.1259016393442623,
|
| 175 |
-
"grad_norm": 0.8192205429077148,
|
| 176 |
-
"learning_rate": 0.0001842076200993926,
|
| 177 |
-
"loss": 0.7307174205780029,
|
| 178 |
-
"step": 240
|
| 179 |
-
},
|
| 180 |
-
{
|
| 181 |
-
"epoch": 0.13114754098360656,
|
| 182 |
-
"grad_norm": 0.6621026396751404,
|
| 183 |
-
"learning_rate": 0.00018310325786858092,
|
| 184 |
-
"loss": 0.7196334362030029,
|
| 185 |
-
"step": 250
|
| 186 |
-
},
|
| 187 |
-
{
|
| 188 |
-
"epoch": 0.13639344262295083,
|
| 189 |
-
"grad_norm": 0.6377759575843811,
|
| 190 |
-
"learning_rate": 0.0001819988956377692,
|
| 191 |
-
"loss": 0.761665153503418,
|
| 192 |
-
"step": 260
|
| 193 |
-
},
|
| 194 |
-
{
|
| 195 |
-
"epoch": 0.14163934426229507,
|
| 196 |
-
"grad_norm": 0.7564328908920288,
|
| 197 |
-
"learning_rate": 0.0001808945334069575,
|
| 198 |
-
"loss": 0.7283723831176758,
|
| 199 |
-
"step": 270
|
| 200 |
-
},
|
| 201 |
-
{
|
| 202 |
-
"epoch": 0.14688524590163934,
|
| 203 |
-
"grad_norm": 0.7781689167022705,
|
| 204 |
-
"learning_rate": 0.00017979017117614578,
|
| 205 |
-
"loss": 0.7663159370422363,
|
| 206 |
-
"step": 280
|
| 207 |
-
},
|
| 208 |
-
{
|
| 209 |
-
"epoch": 0.1521311475409836,
|
| 210 |
-
"grad_norm": 0.7704823017120361,
|
| 211 |
-
"learning_rate": 0.0001786858089453341,
|
| 212 |
-
"loss": 0.7273806095123291,
|
| 213 |
-
"step": 290
|
| 214 |
-
},
|
| 215 |
-
{
|
| 216 |
-
"epoch": 0.15737704918032788,
|
| 217 |
-
"grad_norm": 0.6168962717056274,
|
| 218 |
-
"learning_rate": 0.00017758144671452237,
|
| 219 |
-
"loss": 0.7601676940917969,
|
| 220 |
-
"step": 300
|
| 221 |
-
},
|
| 222 |
-
{
|
| 223 |
-
"epoch": 0.16262295081967212,
|
| 224 |
-
"grad_norm": 0.747313380241394,
|
| 225 |
-
"learning_rate": 0.0001764770844837107,
|
| 226 |
-
"loss": 0.7394940853118896,
|
| 227 |
-
"step": 310
|
| 228 |
-
},
|
| 229 |
-
{
|
| 230 |
-
"epoch": 0.1678688524590164,
|
| 231 |
-
"grad_norm": 0.6736217737197876,
|
| 232 |
-
"learning_rate": 0.00017537272225289895,
|
| 233 |
-
"loss": 0.674213457107544,
|
| 234 |
-
"step": 320
|
| 235 |
-
},
|
| 236 |
-
{
|
| 237 |
-
"epoch": 0.17311475409836066,
|
| 238 |
-
"grad_norm": 0.7672621011734009,
|
| 239 |
-
"learning_rate": 0.00017426836002208728,
|
| 240 |
-
"loss": 0.6635504722595215,
|
| 241 |
-
"step": 330
|
| 242 |
-
},
|
| 243 |
-
{
|
| 244 |
-
"epoch": 0.17836065573770493,
|
| 245 |
-
"grad_norm": 0.6182668209075928,
|
| 246 |
-
"learning_rate": 0.00017316399779127554,
|
| 247 |
-
"loss": 0.6947505950927735,
|
| 248 |
-
"step": 340
|
| 249 |
-
},
|
| 250 |
-
{
|
| 251 |
-
"epoch": 0.18360655737704917,
|
| 252 |
-
"grad_norm": 0.6231220960617065,
|
| 253 |
-
"learning_rate": 0.00017205963556046384,
|
| 254 |
-
"loss": 0.7503675937652587,
|
| 255 |
-
"step": 350
|
| 256 |
-
},
|
| 257 |
-
{
|
| 258 |
-
"epoch": 0.18885245901639344,
|
| 259 |
-
"grad_norm": 0.7776580452919006,
|
| 260 |
-
"learning_rate": 0.00017095527332965213,
|
| 261 |
-
"loss": 0.7351272106170654,
|
| 262 |
-
"step": 360
|
| 263 |
-
},
|
| 264 |
-
{
|
| 265 |
-
"epoch": 0.1940983606557377,
|
| 266 |
-
"grad_norm": 0.7008368968963623,
|
| 267 |
-
"learning_rate": 0.00016985091109884042,
|
| 268 |
-
"loss": 0.6939175605773926,
|
| 269 |
-
"step": 370
|
| 270 |
-
},
|
| 271 |
-
{
|
| 272 |
-
"epoch": 0.19934426229508198,
|
| 273 |
-
"grad_norm": 0.7504845261573792,
|
| 274 |
-
"learning_rate": 0.00016874654886802872,
|
| 275 |
-
"loss": 0.7262158393859863,
|
| 276 |
-
"step": 380
|
| 277 |
-
},
|
| 278 |
-
{
|
| 279 |
-
"epoch": 0.20459016393442622,
|
| 280 |
-
"grad_norm": 0.6740432977676392,
|
| 281 |
-
"learning_rate": 0.000167642186637217,
|
| 282 |
-
"loss": 0.7514342308044434,
|
| 283 |
-
"step": 390
|
| 284 |
-
},
|
| 285 |
-
{
|
| 286 |
-
"epoch": 0.2098360655737705,
|
| 287 |
-
"grad_norm": 0.6686267256736755,
|
| 288 |
-
"learning_rate": 0.0001665378244064053,
|
| 289 |
-
"loss": 0.6774350166320801,
|
| 290 |
-
"step": 400
|
| 291 |
-
},
|
| 292 |
-
{
|
| 293 |
-
"epoch": 0.21508196721311476,
|
| 294 |
-
"grad_norm": 0.5805259943008423,
|
| 295 |
-
"learning_rate": 0.0001654334621755936,
|
| 296 |
-
"loss": 0.6829588413238525,
|
| 297 |
-
"step": 410
|
| 298 |
-
},
|
| 299 |
-
{
|
| 300 |
-
"epoch": 0.22032786885245903,
|
| 301 |
-
"grad_norm": 0.6207637786865234,
|
| 302 |
-
"learning_rate": 0.0001643290999447819,
|
| 303 |
-
"loss": 0.6976052761077881,
|
| 304 |
-
"step": 420
|
| 305 |
-
},
|
| 306 |
-
{
|
| 307 |
-
"epoch": 0.22557377049180327,
|
| 308 |
-
"grad_norm": 1.0807863473892212,
|
| 309 |
-
"learning_rate": 0.0001632247377139702,
|
| 310 |
-
"loss": 0.7167800426483154,
|
| 311 |
-
"step": 430
|
| 312 |
-
},
|
| 313 |
-
{
|
| 314 |
-
"epoch": 0.23081967213114754,
|
| 315 |
-
"grad_norm": 0.6717860698699951,
|
| 316 |
-
"learning_rate": 0.00016212037548315848,
|
| 317 |
-
"loss": 0.701947021484375,
|
| 318 |
-
"step": 440
|
| 319 |
-
},
|
| 320 |
-
{
|
| 321 |
-
"epoch": 0.2360655737704918,
|
| 322 |
-
"grad_norm": 0.6229609847068787,
|
| 323 |
-
"learning_rate": 0.00016101601325234678,
|
| 324 |
-
"loss": 0.6681315898895264,
|
| 325 |
-
"step": 450
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.24131147540983608,
|
| 329 |
-
"grad_norm": 0.8894275426864624,
|
| 330 |
-
"learning_rate": 0.00015991165102153507,
|
| 331 |
-
"loss": 0.7274843692779541,
|
| 332 |
-
"step": 460
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 0.24655737704918032,
|
| 336 |
-
"grad_norm": 0.8598907589912415,
|
| 337 |
-
"learning_rate": 0.00015880728879072336,
|
| 338 |
-
"loss": 0.6865033149719239,
|
| 339 |
-
"step": 470
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.2518032786885246,
|
| 343 |
-
"grad_norm": 1.4645612239837646,
|
| 344 |
-
"learning_rate": 0.00015770292655991166,
|
| 345 |
-
"loss": 0.6898164749145508,
|
| 346 |
-
"step": 480
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 0.25704918032786883,
|
| 350 |
-
"grad_norm": 0.7187657952308655,
|
| 351 |
-
"learning_rate": 0.00015659856432909995,
|
| 352 |
-
"loss": 0.7331682682037354,
|
| 353 |
-
"step": 490
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 0.26229508196721313,
|
| 357 |
-
"grad_norm": 0.6352348327636719,
|
| 358 |
-
"learning_rate": 0.00015549420209828824,
|
| 359 |
-
"loss": 0.6853579998016357,
|
| 360 |
-
"step": 500
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 0.26229508196721313,
|
| 364 |
-
"eval_loss": 0.6940423250198364,
|
| 365 |
-
"eval_runtime": 327.5265,
|
| 366 |
-
"eval_samples_per_second": 34.443,
|
| 367 |
-
"eval_steps_per_second": 8.613,
|
| 368 |
-
"step": 500
|
| 369 |
-
}
|
| 370 |
-
],
|
| 371 |
-
"logging_steps": 10,
|
| 372 |
-
"max_steps": 1907,
|
| 373 |
-
"num_input_tokens_seen": 0,
|
| 374 |
-
"num_train_epochs": 1,
|
| 375 |
-
"save_steps": 500,
|
| 376 |
-
"stateful_callbacks": {
|
| 377 |
-
"TrainerControl": {
|
| 378 |
-
"args": {
|
| 379 |
-
"should_epoch_stop": false,
|
| 380 |
-
"should_evaluate": false,
|
| 381 |
-
"should_log": false,
|
| 382 |
-
"should_save": true,
|
| 383 |
-
"should_training_stop": false
|
| 384 |
-
},
|
| 385 |
-
"attributes": {}
|
| 386 |
-
}
|
| 387 |
-
},
|
| 388 |
-
"total_flos": 2.8628493340644864e+16,
|
| 389 |
-
"train_batch_size": 2,
|
| 390 |
-
"trial_name": null,
|
| 391 |
-
"trial_params": null
|
| 392 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
gemma-3-1b-it/tb/events.out.tfevents.1789955877.fed349279e2c.8616.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bbf98694028a632a474f42a31625599b4e945ee3a7862cf612d5ab0b3fe0edda
|
| 3 |
+
size 47854
|
gemma-3-1b-it/test_eval/metrics.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
-
"wer": 0.
|
| 3 |
"exact_match": 0.0,
|
| 4 |
-
"hallucination_rate": 0.
|
| 5 |
"n_examples": 128
|
| 6 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"wer": 0.8090913703008472,
|
| 3 |
"exact_match": 0.0,
|
| 4 |
+
"hallucination_rate": 0.1953125,
|
| 5 |
"n_examples": 128
|
| 6 |
}
|
gemma-3-1b-it/test_eval/predictions.jsonl
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|