Training in progress, step 300, checkpoint
Browse files- last-checkpoint/adapter_model.safetensors +1 -1
- last-checkpoint/global_step300/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step300/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step300/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step300/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step300/zero_pp_rank_0_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step300/zero_pp_rank_1_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step300/zero_pp_rank_2_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step300/zero_pp_rank_3_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/latest +1 -1
- last-checkpoint/rng_state_0.pth +1 -1
- last-checkpoint/rng_state_1.pth +1 -1
- last-checkpoint/rng_state_2.pth +1 -1
- last-checkpoint/rng_state_3.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +727 -3
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 13648688
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3089ee5752591a99a34512cd610e5fd1c9cbde6f2e5e96052322709db58f3c20
|
| 3 |
size 13648688
|
last-checkpoint/global_step300/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f71c9991f0a3e0cd1179cd2e3874f7b658401de9a5504be3d21b724ba049377d
|
| 3 |
+
size 20450800
|
last-checkpoint/global_step300/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ca1c1e21cddf824e4855926726edc0d137e7bf67de7cf8919d8065cbb405d7c8
|
| 3 |
+
size 20450800
|
last-checkpoint/global_step300/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fb97391dd585cf53fa22f1b33502d8c0078861881bc64091e47c722648ffa9b9
|
| 3 |
+
size 20450800
|
last-checkpoint/global_step300/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:413459590411e0e91e925d5765cb8fff050ffd31b47410bbdaf19c166851ed60
|
| 3 |
+
size 20450800
|
last-checkpoint/global_step300/zero_pp_rank_0_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f10dcf1fc0a43f6046928990df268bbc1f546b31d3568428be3f358423cb0951
|
| 3 |
+
size 152238
|
last-checkpoint/global_step300/zero_pp_rank_1_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ca6a6f9616d9eb1bef7a71132384d000d3064294e87f3146629a6a6822eddcfa
|
| 3 |
+
size 152238
|
last-checkpoint/global_step300/zero_pp_rank_2_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e68119fc8ce07b421ca4fab0993069c3e9775c9c9d11bd16ec80b5f4bf8c9848
|
| 3 |
+
size 152238
|
last-checkpoint/global_step300/zero_pp_rank_3_mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b76d5ab8fb6578710d191973f5159a7cef6a0564d586d1346cf7637a81320c38
|
| 3 |
+
size 152238
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step300
|
last-checkpoint/rng_state_0.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:680c0f067459bb4efdac849ce093e2226bf3c2332330a52eb68acec721890eea
|
| 3 |
size 14960
|
last-checkpoint/rng_state_1.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a8261bb0773aaaaaf837917ac2d74751a1b07817c980444e7109f977082d4d80
|
| 3 |
size 14960
|
last-checkpoint/rng_state_2.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7ba4d6439beb986cf1f95fd682e03fa5844ac212a382301bdd1a868bcc67c311
|
| 3 |
size 14960
|
last-checkpoint/rng_state_3.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14960
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c17d861750a27e832ec9cf9a840f42cdc22319da36842441a78feca72092cef2
|
| 3 |
size 14960
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5c0f31e0dd8f0cd1067395334590c80ca29a3a9a42118ffbf479961406c7bb0b
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 40,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1479,6 +1479,730 @@
|
|
| 1479 |
"eval_samples_per_second": 2.12,
|
| 1480 |
"eval_steps_per_second": 0.17,
|
| 1481 |
"step": 200
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1482 |
}
|
| 1483 |
],
|
| 1484 |
"logging_steps": 1,
|
|
@@ -1498,7 +2222,7 @@
|
|
| 1498 |
"attributes": {}
|
| 1499 |
}
|
| 1500 |
},
|
| 1501 |
-
"total_flos":
|
| 1502 |
"train_batch_size": 4,
|
| 1503 |
"trial_name": null,
|
| 1504 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.3688902551490931,
|
| 5 |
"eval_steps": 40,
|
| 6 |
+
"global_step": 300,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1479 |
"eval_samples_per_second": 2.12,
|
| 1480 |
"eval_steps_per_second": 0.17,
|
| 1481 |
"step": 200
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 0.2471564709498924,
|
| 1485 |
+
"grad_norm": 0.26939804480510793,
|
| 1486 |
+
"learning_rate": 1.872049408143808e-05,
|
| 1487 |
+
"loss": 0.1785,
|
| 1488 |
+
"step": 201
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 0.24838610513372272,
|
| 1492 |
+
"grad_norm": 0.3016410055860768,
|
| 1493 |
+
"learning_rate": 1.8699380076440242e-05,
|
| 1494 |
+
"loss": 0.2045,
|
| 1495 |
+
"step": 202
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 0.24961573931755301,
|
| 1499 |
+
"grad_norm": 0.3469258635915736,
|
| 1500 |
+
"learning_rate": 1.8678105394844114e-05,
|
| 1501 |
+
"loss": 0.2027,
|
| 1502 |
+
"step": 203
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 0.25084537350138336,
|
| 1506 |
+
"grad_norm": 0.32607531852363275,
|
| 1507 |
+
"learning_rate": 1.8656670429590745e-05,
|
| 1508 |
+
"loss": 0.1564,
|
| 1509 |
+
"step": 204
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 0.2520750076852136,
|
| 1513 |
+
"grad_norm": 0.2760811969255309,
|
| 1514 |
+
"learning_rate": 1.8635075576581587e-05,
|
| 1515 |
+
"loss": 0.1723,
|
| 1516 |
+
"step": 205
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 0.25330464186904394,
|
| 1520 |
+
"grad_norm": 0.25212411563493814,
|
| 1521 |
+
"learning_rate": 1.861332123467122e-05,
|
| 1522 |
+
"loss": 0.2283,
|
| 1523 |
+
"step": 206
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 0.25453427605287426,
|
| 1527 |
+
"grad_norm": 0.3448353672449361,
|
| 1528 |
+
"learning_rate": 1.859140780565996e-05,
|
| 1529 |
+
"loss": 0.2186,
|
| 1530 |
+
"step": 207
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 0.2557639102367046,
|
| 1534 |
+
"grad_norm": 0.35723591175806685,
|
| 1535 |
+
"learning_rate": 1.856933569428644e-05,
|
| 1536 |
+
"loss": 0.2346,
|
| 1537 |
+
"step": 208
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 0.2569935444205349,
|
| 1541 |
+
"grad_norm": 0.2721994057338676,
|
| 1542 |
+
"learning_rate": 1.8547105308220142e-05,
|
| 1543 |
+
"loss": 0.2408,
|
| 1544 |
+
"step": 209
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 0.2582231786043652,
|
| 1548 |
+
"grad_norm": 0.2946335121583992,
|
| 1549 |
+
"learning_rate": 1.852471705805387e-05,
|
| 1550 |
+
"loss": 0.179,
|
| 1551 |
+
"step": 210
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 0.2594528127881955,
|
| 1555 |
+
"grad_norm": 0.44468973852329186,
|
| 1556 |
+
"learning_rate": 1.8502171357296144e-05,
|
| 1557 |
+
"loss": 0.2007,
|
| 1558 |
+
"step": 211
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 0.2606824469720258,
|
| 1562 |
+
"grad_norm": 0.4480447012946234,
|
| 1563 |
+
"learning_rate": 1.84794686223636e-05,
|
| 1564 |
+
"loss": 0.2752,
|
| 1565 |
+
"step": 212
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 0.2619120811558561,
|
| 1569 |
+
"grad_norm": 0.28800115930201,
|
| 1570 |
+
"learning_rate": 1.8456609272573268e-05,
|
| 1571 |
+
"loss": 0.2106,
|
| 1572 |
+
"step": 213
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 0.26314171533968644,
|
| 1576 |
+
"grad_norm": 0.4124378870162344,
|
| 1577 |
+
"learning_rate": 1.8433593730134835e-05,
|
| 1578 |
+
"loss": 0.2648,
|
| 1579 |
+
"step": 214
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 0.26437134952351676,
|
| 1583 |
+
"grad_norm": 0.26984478766637626,
|
| 1584 |
+
"learning_rate": 1.841042242014285e-05,
|
| 1585 |
+
"loss": 0.1578,
|
| 1586 |
+
"step": 215
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 0.2656009837073471,
|
| 1590 |
+
"grad_norm": 0.4271224156022197,
|
| 1591 |
+
"learning_rate": 1.838709577056888e-05,
|
| 1592 |
+
"loss": 0.241,
|
| 1593 |
+
"step": 216
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 0.2668306178911774,
|
| 1597 |
+
"grad_norm": 0.32441622820215904,
|
| 1598 |
+
"learning_rate": 1.8363614212253585e-05,
|
| 1599 |
+
"loss": 0.1615,
|
| 1600 |
+
"step": 217
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 0.26806025207500767,
|
| 1604 |
+
"grad_norm": 0.35733117717955937,
|
| 1605 |
+
"learning_rate": 1.833997817889878e-05,
|
| 1606 |
+
"loss": 0.234,
|
| 1607 |
+
"step": 218
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 0.269289886258838,
|
| 1611 |
+
"grad_norm": 0.3419406717090276,
|
| 1612 |
+
"learning_rate": 1.8316188107059418e-05,
|
| 1613 |
+
"loss": 0.2538,
|
| 1614 |
+
"step": 219
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 0.2705195204426683,
|
| 1618 |
+
"grad_norm": 0.31666227946792574,
|
| 1619 |
+
"learning_rate": 1.8292244436135517e-05,
|
| 1620 |
+
"loss": 0.2518,
|
| 1621 |
+
"step": 220
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 0.2717491546264986,
|
| 1625 |
+
"grad_norm": 0.25628519133100447,
|
| 1626 |
+
"learning_rate": 1.8268147608364068e-05,
|
| 1627 |
+
"loss": 0.1488,
|
| 1628 |
+
"step": 221
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 0.27297878881032894,
|
| 1632 |
+
"grad_norm": 0.40851754881119196,
|
| 1633 |
+
"learning_rate": 1.8243898068810833e-05,
|
| 1634 |
+
"loss": 0.2662,
|
| 1635 |
+
"step": 222
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 0.27420842299415926,
|
| 1639 |
+
"grad_norm": 0.38439452767162985,
|
| 1640 |
+
"learning_rate": 1.8219496265362164e-05,
|
| 1641 |
+
"loss": 0.2033,
|
| 1642 |
+
"step": 223
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 0.2754380571779895,
|
| 1646 |
+
"grad_norm": 0.39575941152172794,
|
| 1647 |
+
"learning_rate": 1.81949426487167e-05,
|
| 1648 |
+
"loss": 0.2481,
|
| 1649 |
+
"step": 224
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 0.27666769136181985,
|
| 1653 |
+
"grad_norm": 0.39720852466696316,
|
| 1654 |
+
"learning_rate": 1.8170237672377046e-05,
|
| 1655 |
+
"loss": 0.2712,
|
| 1656 |
+
"step": 225
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 0.27789732554565016,
|
| 1660 |
+
"grad_norm": 0.56556586568128,
|
| 1661 |
+
"learning_rate": 1.814538179264142e-05,
|
| 1662 |
+
"loss": 0.2856,
|
| 1663 |
+
"step": 226
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 0.2791269597294805,
|
| 1667 |
+
"grad_norm": 0.4627206566124822,
|
| 1668 |
+
"learning_rate": 1.81203754685952e-05,
|
| 1669 |
+
"loss": 0.2353,
|
| 1670 |
+
"step": 227
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 0.2803565939133108,
|
| 1674 |
+
"grad_norm": 0.2862726313485586,
|
| 1675 |
+
"learning_rate": 1.8095219162102453e-05,
|
| 1676 |
+
"loss": 0.1773,
|
| 1677 |
+
"step": 228
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 0.2815862280971411,
|
| 1681 |
+
"grad_norm": 0.36128900613055054,
|
| 1682 |
+
"learning_rate": 1.8069913337797414e-05,
|
| 1683 |
+
"loss": 0.2265,
|
| 1684 |
+
"step": 229
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 0.2828158622809714,
|
| 1688 |
+
"grad_norm": 0.3598391415884193,
|
| 1689 |
+
"learning_rate": 1.804445846307588e-05,
|
| 1690 |
+
"loss": 0.2167,
|
| 1691 |
+
"step": 230
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 0.2840454964648017,
|
| 1695 |
+
"grad_norm": 0.3038583906156817,
|
| 1696 |
+
"learning_rate": 1.801885500808661e-05,
|
| 1697 |
+
"loss": 0.1929,
|
| 1698 |
+
"step": 231
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 0.285275130648632,
|
| 1702 |
+
"grad_norm": 0.2684590976129711,
|
| 1703 |
+
"learning_rate": 1.7993103445722615e-05,
|
| 1704 |
+
"loss": 0.1955,
|
| 1705 |
+
"step": 232
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 0.28650476483246234,
|
| 1709 |
+
"grad_norm": 0.3650334815395792,
|
| 1710 |
+
"learning_rate": 1.7967204251612432e-05,
|
| 1711 |
+
"loss": 0.2021,
|
| 1712 |
+
"step": 233
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 0.28773439901629266,
|
| 1716 |
+
"grad_norm": 0.3529582774863101,
|
| 1717 |
+
"learning_rate": 1.7941157904111346e-05,
|
| 1718 |
+
"loss": 0.2396,
|
| 1719 |
+
"step": 234
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 0.288964033200123,
|
| 1723 |
+
"grad_norm": 0.24658124615059435,
|
| 1724 |
+
"learning_rate": 1.7914964884292543e-05,
|
| 1725 |
+
"loss": 0.1841,
|
| 1726 |
+
"step": 235
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 0.29019366738395325,
|
| 1730 |
+
"grad_norm": 0.3897045768921742,
|
| 1731 |
+
"learning_rate": 1.7888625675938237e-05,
|
| 1732 |
+
"loss": 0.2233,
|
| 1733 |
+
"step": 236
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 0.29142330156778357,
|
| 1737 |
+
"grad_norm": 0.24436873250999808,
|
| 1738 |
+
"learning_rate": 1.7862140765530718e-05,
|
| 1739 |
+
"loss": 0.1587,
|
| 1740 |
+
"step": 237
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 0.2926529357516139,
|
| 1744 |
+
"grad_norm": 0.4324084079497792,
|
| 1745 |
+
"learning_rate": 1.783551064224339e-05,
|
| 1746 |
+
"loss": 0.1914,
|
| 1747 |
+
"step": 238
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 0.2938825699354442,
|
| 1751 |
+
"grad_norm": 0.31640333118588265,
|
| 1752 |
+
"learning_rate": 1.7808735797931715e-05,
|
| 1753 |
+
"loss": 0.1512,
|
| 1754 |
+
"step": 239
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 0.2951122041192745,
|
| 1758 |
+
"grad_norm": 0.34244615386284427,
|
| 1759 |
+
"learning_rate": 1.7781816727124138e-05,
|
| 1760 |
+
"loss": 0.2004,
|
| 1761 |
+
"step": 240
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 0.2951122041192745,
|
| 1765 |
+
"eval_accuracy": 0.7807486631016043,
|
| 1766 |
+
"eval_f1": 0.4225352112676056,
|
| 1767 |
+
"eval_loss": 0.4479687511920929,
|
| 1768 |
+
"eval_precision": 0.7142857142857143,
|
| 1769 |
+
"eval_recall": 0.3,
|
| 1770 |
+
"eval_runtime": 22.697,
|
| 1771 |
+
"eval_samples_per_second": 2.203,
|
| 1772 |
+
"eval_steps_per_second": 0.176,
|
| 1773 |
+
"step": 240
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 0.29634183830310484,
|
| 1777 |
+
"grad_norm": 0.24395125168807116,
|
| 1778 |
+
"learning_rate": 1.7754753927012955e-05,
|
| 1779 |
+
"loss": 0.1768,
|
| 1780 |
+
"step": 241
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.29757147248693516,
|
| 1784 |
+
"grad_norm": 0.2464638831816697,
|
| 1785 |
+
"learning_rate": 1.7727547897445117e-05,
|
| 1786 |
+
"loss": 0.1461,
|
| 1787 |
+
"step": 242
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 0.2988011066707654,
|
| 1791 |
+
"grad_norm": 0.33968511169569393,
|
| 1792 |
+
"learning_rate": 1.770019914091302e-05,
|
| 1793 |
+
"loss": 0.161,
|
| 1794 |
+
"step": 243
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 0.30003074085459575,
|
| 1798 |
+
"grad_norm": 0.6034876950604299,
|
| 1799 |
+
"learning_rate": 1.76727081625452e-05,
|
| 1800 |
+
"loss": 0.2023,
|
| 1801 |
+
"step": 244
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.30126037503842606,
|
| 1805 |
+
"grad_norm": 0.4452183735534574,
|
| 1806 |
+
"learning_rate": 1.7645075470097024e-05,
|
| 1807 |
+
"loss": 0.2207,
|
| 1808 |
+
"step": 245
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.3024900092222564,
|
| 1812 |
+
"grad_norm": 0.29139249592342986,
|
| 1813 |
+
"learning_rate": 1.7617301573941296e-05,
|
| 1814 |
+
"loss": 0.1763,
|
| 1815 |
+
"step": 246
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 0.3037196434060867,
|
| 1819 |
+
"grad_norm": 0.33614977158297427,
|
| 1820 |
+
"learning_rate": 1.758938698705884e-05,
|
| 1821 |
+
"loss": 0.2381,
|
| 1822 |
+
"step": 247
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 0.304949277589917,
|
| 1826 |
+
"grad_norm": 0.45585634062118247,
|
| 1827 |
+
"learning_rate": 1.7561332225029022e-05,
|
| 1828 |
+
"loss": 0.2215,
|
| 1829 |
+
"step": 248
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 0.3061789117737473,
|
| 1833 |
+
"grad_norm": 0.3889096135056135,
|
| 1834 |
+
"learning_rate": 1.7533137806020226e-05,
|
| 1835 |
+
"loss": 0.2512,
|
| 1836 |
+
"step": 249
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.3074085459575776,
|
| 1840 |
+
"grad_norm": 0.3698201068005034,
|
| 1841 |
+
"learning_rate": 1.7504804250780292e-05,
|
| 1842 |
+
"loss": 0.1962,
|
| 1843 |
+
"step": 250
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.3086381801414079,
|
| 1847 |
+
"grad_norm": 0.23954623770426672,
|
| 1848 |
+
"learning_rate": 1.747633208262688e-05,
|
| 1849 |
+
"loss": 0.1669,
|
| 1850 |
+
"step": 251
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.30986781432523824,
|
| 1854 |
+
"grad_norm": 0.3069850649077184,
|
| 1855 |
+
"learning_rate": 1.744772182743782e-05,
|
| 1856 |
+
"loss": 0.173,
|
| 1857 |
+
"step": 252
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.31109744850906856,
|
| 1861 |
+
"grad_norm": 0.5346208398414731,
|
| 1862 |
+
"learning_rate": 1.74189740136414e-05,
|
| 1863 |
+
"loss": 0.2977,
|
| 1864 |
+
"step": 253
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.3123270826928989,
|
| 1868 |
+
"grad_norm": 0.4383899410481399,
|
| 1869 |
+
"learning_rate": 1.7390089172206594e-05,
|
| 1870 |
+
"loss": 0.1867,
|
| 1871 |
+
"step": 254
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.31355671687672915,
|
| 1875 |
+
"grad_norm": 0.34336119858301023,
|
| 1876 |
+
"learning_rate": 1.736106783663326e-05,
|
| 1877 |
+
"loss": 0.2143,
|
| 1878 |
+
"step": 255
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.31478635106055947,
|
| 1882 |
+
"grad_norm": 0.2605528903698455,
|
| 1883 |
+
"learning_rate": 1.7331910542942298e-05,
|
| 1884 |
+
"loss": 0.2061,
|
| 1885 |
+
"step": 256
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.3160159852443898,
|
| 1889 |
+
"grad_norm": 0.29064553728173864,
|
| 1890 |
+
"learning_rate": 1.7302617829665725e-05,
|
| 1891 |
+
"loss": 0.1888,
|
| 1892 |
+
"step": 257
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.3172456194282201,
|
| 1896 |
+
"grad_norm": 0.42180400278336094,
|
| 1897 |
+
"learning_rate": 1.7273190237836757e-05,
|
| 1898 |
+
"loss": 0.1727,
|
| 1899 |
+
"step": 258
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.3184752536120504,
|
| 1903 |
+
"grad_norm": 0.5419289309694173,
|
| 1904 |
+
"learning_rate": 1.7243628310979793e-05,
|
| 1905 |
+
"loss": 0.2215,
|
| 1906 |
+
"step": 259
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.31970488779588074,
|
| 1910 |
+
"grad_norm": 0.5826869965037216,
|
| 1911 |
+
"learning_rate": 1.7213932595100384e-05,
|
| 1912 |
+
"loss": 0.2394,
|
| 1913 |
+
"step": 260
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.32093452197971106,
|
| 1917 |
+
"grad_norm": 0.35549156755432715,
|
| 1918 |
+
"learning_rate": 1.7184103638675157e-05,
|
| 1919 |
+
"loss": 0.2212,
|
| 1920 |
+
"step": 261
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.3221641561635413,
|
| 1924 |
+
"grad_norm": 0.403587329803515,
|
| 1925 |
+
"learning_rate": 1.715414199264168e-05,
|
| 1926 |
+
"loss": 0.1709,
|
| 1927 |
+
"step": 262
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 0.32339379034737165,
|
| 1931 |
+
"grad_norm": 0.568711959322274,
|
| 1932 |
+
"learning_rate": 1.7124048210388268e-05,
|
| 1933 |
+
"loss": 0.1972,
|
| 1934 |
+
"step": 263
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 0.32462342453120197,
|
| 1938 |
+
"grad_norm": 0.23810294888269123,
|
| 1939 |
+
"learning_rate": 1.709382284774379e-05,
|
| 1940 |
+
"loss": 0.1846,
|
| 1941 |
+
"step": 264
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 0.3258530587150323,
|
| 1945 |
+
"grad_norm": 0.3902059118868234,
|
| 1946 |
+
"learning_rate": 1.706346646296739e-05,
|
| 1947 |
+
"loss": 0.2197,
|
| 1948 |
+
"step": 265
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 0.3270826928988626,
|
| 1952 |
+
"grad_norm": 0.30966956204734636,
|
| 1953 |
+
"learning_rate": 1.7032979616738167e-05,
|
| 1954 |
+
"loss": 0.1728,
|
| 1955 |
+
"step": 266
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 0.3283123270826929,
|
| 1959 |
+
"grad_norm": 0.3364767329741329,
|
| 1960 |
+
"learning_rate": 1.7002362872144843e-05,
|
| 1961 |
+
"loss": 0.236,
|
| 1962 |
+
"step": 267
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 0.3295419612665232,
|
| 1966 |
+
"grad_norm": 0.2575556372712217,
|
| 1967 |
+
"learning_rate": 1.697161679467534e-05,
|
| 1968 |
+
"loss": 0.2166,
|
| 1969 |
+
"step": 268
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 0.3307715954503535,
|
| 1973 |
+
"grad_norm": 0.6139887685526306,
|
| 1974 |
+
"learning_rate": 1.6940741952206342e-05,
|
| 1975 |
+
"loss": 0.217,
|
| 1976 |
+
"step": 269
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 0.3320012296341838,
|
| 1980 |
+
"grad_norm": 0.3192026684409325,
|
| 1981 |
+
"learning_rate": 1.6909738914992812e-05,
|
| 1982 |
+
"loss": 0.228,
|
| 1983 |
+
"step": 270
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 0.33323086381801414,
|
| 1987 |
+
"grad_norm": 0.3191098195045691,
|
| 1988 |
+
"learning_rate": 1.6878608255657457e-05,
|
| 1989 |
+
"loss": 0.2148,
|
| 1990 |
+
"step": 271
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 0.33446049800184446,
|
| 1994 |
+
"grad_norm": 0.3876596561870598,
|
| 1995 |
+
"learning_rate": 1.6847350549180148e-05,
|
| 1996 |
+
"loss": 0.2191,
|
| 1997 |
+
"step": 272
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 0.3356901321856748,
|
| 2001 |
+
"grad_norm": 0.38903679735549906,
|
| 2002 |
+
"learning_rate": 1.6815966372887305e-05,
|
| 2003 |
+
"loss": 0.2205,
|
| 2004 |
+
"step": 273
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 0.33691976636950505,
|
| 2008 |
+
"grad_norm": 0.38626350258480885,
|
| 2009 |
+
"learning_rate": 1.6784456306441234e-05,
|
| 2010 |
+
"loss": 0.2672,
|
| 2011 |
+
"step": 274
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.33814940055333537,
|
| 2015 |
+
"grad_norm": 0.3297053884257871,
|
| 2016 |
+
"learning_rate": 1.675282093182941e-05,
|
| 2017 |
+
"loss": 0.1686,
|
| 2018 |
+
"step": 275
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 0.3393790347371657,
|
| 2022 |
+
"grad_norm": 0.4263354582166456,
|
| 2023 |
+
"learning_rate": 1.672106083335374e-05,
|
| 2024 |
+
"loss": 0.1873,
|
| 2025 |
+
"step": 276
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.340608668920996,
|
| 2029 |
+
"grad_norm": 0.2796799138264646,
|
| 2030 |
+
"learning_rate": 1.6689176597619773e-05,
|
| 2031 |
+
"loss": 0.1861,
|
| 2032 |
+
"step": 277
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.3418383031048263,
|
| 2036 |
+
"grad_norm": 0.2741188381316141,
|
| 2037 |
+
"learning_rate": 1.6657168813525855e-05,
|
| 2038 |
+
"loss": 0.2212,
|
| 2039 |
+
"step": 278
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 0.34306793728865664,
|
| 2043 |
+
"grad_norm": 0.3392975873746286,
|
| 2044 |
+
"learning_rate": 1.662503807225225e-05,
|
| 2045 |
+
"loss": 0.2087,
|
| 2046 |
+
"step": 279
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 0.3442975714724869,
|
| 2050 |
+
"grad_norm": 0.3420260746414455,
|
| 2051 |
+
"learning_rate": 1.659278496725024e-05,
|
| 2052 |
+
"loss": 0.2241,
|
| 2053 |
+
"step": 280
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 0.3442975714724869,
|
| 2057 |
+
"eval_accuracy": 0.7807486631016043,
|
| 2058 |
+
"eval_f1": 0.4225352112676056,
|
| 2059 |
+
"eval_loss": 0.4449218809604645,
|
| 2060 |
+
"eval_precision": 0.7142857142857143,
|
| 2061 |
+
"eval_recall": 0.3,
|
| 2062 |
+
"eval_runtime": 22.916,
|
| 2063 |
+
"eval_samples_per_second": 2.182,
|
| 2064 |
+
"eval_steps_per_second": 0.175,
|
| 2065 |
+
"step": 280
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 0.3455272056563172,
|
| 2069 |
+
"grad_norm": 0.34182478583361364,
|
| 2070 |
+
"learning_rate": 1.6560410094231144e-05,
|
| 2071 |
+
"loss": 0.2257,
|
| 2072 |
+
"step": 281
|
| 2073 |
+
},
|
| 2074 |
+
{
|
| 2075 |
+
"epoch": 0.34675683984014755,
|
| 2076 |
+
"grad_norm": 0.3981831055380401,
|
| 2077 |
+
"learning_rate": 1.6527914051155328e-05,
|
| 2078 |
+
"loss": 0.2593,
|
| 2079 |
+
"step": 282
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 0.34798647402397787,
|
| 2083 |
+
"grad_norm": 0.3376266386557847,
|
| 2084 |
+
"learning_rate": 1.6495297438221145e-05,
|
| 2085 |
+
"loss": 0.1899,
|
| 2086 |
+
"step": 283
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 0.3492161082078082,
|
| 2090 |
+
"grad_norm": 0.3924075494987157,
|
| 2091 |
+
"learning_rate": 1.6462560857853876e-05,
|
| 2092 |
+
"loss": 0.2032,
|
| 2093 |
+
"step": 284
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 0.3504457423916385,
|
| 2097 |
+
"grad_norm": 0.5277347648864814,
|
| 2098 |
+
"learning_rate": 1.6429704914694573e-05,
|
| 2099 |
+
"loss": 0.1546,
|
| 2100 |
+
"step": 285
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 0.3516753765754688,
|
| 2104 |
+
"grad_norm": 0.32345022677167856,
|
| 2105 |
+
"learning_rate": 1.6396730215588913e-05,
|
| 2106 |
+
"loss": 0.2559,
|
| 2107 |
+
"step": 286
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"epoch": 0.3529050107592991,
|
| 2111 |
+
"grad_norm": 0.44254976480188857,
|
| 2112 |
+
"learning_rate": 1.6363637369575984e-05,
|
| 2113 |
+
"loss": 0.2794,
|
| 2114 |
+
"step": 287
|
| 2115 |
+
},
|
| 2116 |
+
{
|
| 2117 |
+
"epoch": 0.3541346449431294,
|
| 2118 |
+
"grad_norm": 0.4548791128214442,
|
| 2119 |
+
"learning_rate": 1.633042698787703e-05,
|
| 2120 |
+
"loss": 0.2879,
|
| 2121 |
+
"step": 288
|
| 2122 |
+
},
|
| 2123 |
+
{
|
| 2124 |
+
"epoch": 0.3553642791269597,
|
| 2125 |
+
"grad_norm": 0.27873749207646975,
|
| 2126 |
+
"learning_rate": 1.6297099683884163e-05,
|
| 2127 |
+
"loss": 0.1636,
|
| 2128 |
+
"step": 289
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 0.35659391331079004,
|
| 2132 |
+
"grad_norm": 0.3258495849943053,
|
| 2133 |
+
"learning_rate": 1.626365607314905e-05,
|
| 2134 |
+
"loss": 0.2007,
|
| 2135 |
+
"step": 290
|
| 2136 |
+
},
|
| 2137 |
+
{
|
| 2138 |
+
"epoch": 0.35782354749462036,
|
| 2139 |
+
"grad_norm": 0.4098308834550629,
|
| 2140 |
+
"learning_rate": 1.6230096773371514e-05,
|
| 2141 |
+
"loss": 0.2787,
|
| 2142 |
+
"step": 291
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 0.3590531816784507,
|
| 2146 |
+
"grad_norm": 0.2722654753472478,
|
| 2147 |
+
"learning_rate": 1.619642240438816e-05,
|
| 2148 |
+
"loss": 0.19,
|
| 2149 |
+
"step": 292
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 0.36028281586228095,
|
| 2153 |
+
"grad_norm": 0.28423296754063243,
|
| 2154 |
+
"learning_rate": 1.616263358816089e-05,
|
| 2155 |
+
"loss": 0.1652,
|
| 2156 |
+
"step": 293
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 0.36151245004611127,
|
| 2160 |
+
"grad_norm": 0.29897633728182227,
|
| 2161 |
+
"learning_rate": 1.612873094876545e-05,
|
| 2162 |
+
"loss": 0.2339,
|
| 2163 |
+
"step": 294
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 0.3627420842299416,
|
| 2167 |
+
"grad_norm": 0.31056546478547936,
|
| 2168 |
+
"learning_rate": 1.6094715112379874e-05,
|
| 2169 |
+
"loss": 0.1838,
|
| 2170 |
+
"step": 295
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 0.3639717184137719,
|
| 2174 |
+
"grad_norm": 0.2931202881800801,
|
| 2175 |
+
"learning_rate": 1.6060586707272943e-05,
|
| 2176 |
+
"loss": 0.1744,
|
| 2177 |
+
"step": 296
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 0.3652013525976022,
|
| 2181 |
+
"grad_norm": 0.37366320482050214,
|
| 2182 |
+
"learning_rate": 1.6026346363792565e-05,
|
| 2183 |
+
"loss": 0.1915,
|
| 2184 |
+
"step": 297
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 0.36643098678143254,
|
| 2188 |
+
"grad_norm": 0.3528926787344543,
|
| 2189 |
+
"learning_rate": 1.599199471435414e-05,
|
| 2190 |
+
"loss": 0.1589,
|
| 2191 |
+
"step": 298
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 0.3676606209652628,
|
| 2195 |
+
"grad_norm": 0.3024625428052122,
|
| 2196 |
+
"learning_rate": 1.5957532393428872e-05,
|
| 2197 |
+
"loss": 0.2285,
|
| 2198 |
+
"step": 299
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 0.3688902551490931,
|
| 2202 |
+
"grad_norm": 0.3986663078423136,
|
| 2203 |
+
"learning_rate": 1.5922960037532057e-05,
|
| 2204 |
+
"loss": 0.281,
|
| 2205 |
+
"step": 300
|
| 2206 |
}
|
| 2207 |
],
|
| 2208 |
"logging_steps": 1,
|
|
|
|
| 2222 |
"attributes": {}
|
| 2223 |
}
|
| 2224 |
},
|
| 2225 |
+
"total_flos": 295277918322688.0,
|
| 2226 |
"train_batch_size": 4,
|
| 2227 |
"trial_name": null,
|
| 2228 |
"trial_params": null
|