Training in progress, epoch 2, checkpoint
Browse files- last-checkpoint/global_step1668/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step1668/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step1668/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step1668/zero_pp_rank_0_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step1668/zero_pp_rank_1_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step1668/zero_pp_rank_2_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/latest +1 -1
- last-checkpoint/model-00001-of-00004.safetensors +1 -1
- last-checkpoint/model-00002-of-00004.safetensors +1 -1
- last-checkpoint/model-00003-of-00004.safetensors +1 -1
- last-checkpoint/model-00004-of-00004.safetensors +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/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +1263 -2
last-checkpoint/global_step1668/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:cf98bffcabab8218507bcba9b08368256926e85c99089424f03b54f5dd9bd564
|
| 3 |
+
size 20308318462
|
last-checkpoint/global_step1668/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:c6966e5c95822174294276e262d2f292b9bfa7cf9e9a2fac124b02c3b7fb1339
|
| 3 |
+
size 20308318462
|
last-checkpoint/global_step1668/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 sha256:64973c0ff25dd503d332535a3c7d43691046968cdd5c0a3bd0ecb94abee6cc74
|
| 3 |
+
size 20308318462
|
last-checkpoint/global_step1668/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 sha256:92fc02ffa815494ea7895effd23c436d7e421c0c2bf2295fb48d6d736e572c37
|
| 3 |
+
size 168021
|
last-checkpoint/global_step1668/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:db2740855c548b88fdd2e6b1edadbbe5d6ba4d389780b6c44459f5c0eff7c91c
|
| 3 |
+
size 168021
|
last-checkpoint/global_step1668/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:a542c036900a4d52c338793db1098af73debb728de0d9d24f627e9ec9374933f
|
| 3 |
+
size 168021
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step1668
|
last-checkpoint/model-00001-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4877660776
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:602525917b9baaf9f4ce832e2e8edf65f80cd28df8d88e88a66c5455ad0b159f
|
| 3 |
size 4877660776
|
last-checkpoint/model-00002-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4932751008
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8f873ecc91c7a3d668d4c239cff7184b6e599c2084aa976701d7ecdcaaafc6b9
|
| 3 |
size 4932751008
|
last-checkpoint/model-00003-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4330865200
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c0d5018d005251b127a1846019c48d630f6d77fa79e87cde98f2bc6400bb7d91
|
| 3 |
size 4330865200
|
last-checkpoint/model-00004-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1089994880
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:03d7b7c489fa4f5c9a58e26ef7976f8463630dab2895a97e6c976a0edf022e0e
|
| 3 |
size 1089994880
|
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 14768
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9698021f2d84167912e7be6ba48d3d2b8d6b20894f23319f36df078c03b33a64
|
| 3 |
size 14768
|
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 14768
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90a140d1d010220b1679bf6e519f8d3d518cb57331e0e7fb30008dc00e427811
|
| 3 |
size 14768
|
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 14768
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:19aef773503e08b43c9bd940d36e298220b8d39900e7bd698b6996ac3625e59e
|
| 3 |
size 14768
|
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:dd0242fe78905f8cb32fe932e8bcb70076d2384705e561f1118d71e5d750d8b4
|
| 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":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -1283,6 +1283,1267 @@
|
|
| 1283 |
"eval_samples_per_second": 20.053,
|
| 1284 |
"eval_steps_per_second": 0.85,
|
| 1285 |
"step": 834
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1286 |
}
|
| 1287 |
],
|
| 1288 |
"logging_steps": 10,
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1668,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 1283 |
"eval_samples_per_second": 20.053,
|
| 1284 |
"eval_steps_per_second": 0.85,
|
| 1285 |
"step": 834
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 1.0071942446043165,
|
| 1289 |
+
"grad_norm": 18.731496491285114,
|
| 1290 |
+
"learning_rate": 3.691692581075078e-07,
|
| 1291 |
+
"logits/chosen": -2.359375,
|
| 1292 |
+
"logits/rejected": -2.375,
|
| 1293 |
+
"logps/chosen": -237.0,
|
| 1294 |
+
"logps/rejected": -272.0,
|
| 1295 |
+
"loss": 0.2766,
|
| 1296 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1297 |
+
"rewards/chosen": -1.9453125,
|
| 1298 |
+
"rewards/margins": 1.984375,
|
| 1299 |
+
"rewards/rejected": -3.921875,
|
| 1300 |
+
"step": 840
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 1.0191846522781776,
|
| 1304 |
+
"grad_norm": 18.062862619034647,
|
| 1305 |
+
"learning_rate": 3.6694802310084405e-07,
|
| 1306 |
+
"logits/chosen": -2.265625,
|
| 1307 |
+
"logits/rejected": -2.296875,
|
| 1308 |
+
"logps/chosen": -226.0,
|
| 1309 |
+
"logps/rejected": -246.0,
|
| 1310 |
+
"loss": 0.2457,
|
| 1311 |
+
"rewards/accuracies": 0.9375,
|
| 1312 |
+
"rewards/chosen": -1.625,
|
| 1313 |
+
"rewards/margins": 1.8046875,
|
| 1314 |
+
"rewards/rejected": -3.4375,
|
| 1315 |
+
"step": 850
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 1.0311750599520384,
|
| 1319 |
+
"grad_norm": 18.43054843437252,
|
| 1320 |
+
"learning_rate": 3.6472678809418033e-07,
|
| 1321 |
+
"logits/chosen": -2.28125,
|
| 1322 |
+
"logits/rejected": -2.34375,
|
| 1323 |
+
"logps/chosen": -234.0,
|
| 1324 |
+
"logps/rejected": -239.0,
|
| 1325 |
+
"loss": 0.2445,
|
| 1326 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 1327 |
+
"rewards/chosen": -1.7734375,
|
| 1328 |
+
"rewards/margins": 1.984375,
|
| 1329 |
+
"rewards/rejected": -3.765625,
|
| 1330 |
+
"step": 860
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 1.0431654676258992,
|
| 1334 |
+
"grad_norm": 14.792270635313406,
|
| 1335 |
+
"learning_rate": 3.625055530875166e-07,
|
| 1336 |
+
"logits/chosen": -2.328125,
|
| 1337 |
+
"logits/rejected": -2.296875,
|
| 1338 |
+
"logps/chosen": -239.0,
|
| 1339 |
+
"logps/rejected": -258.0,
|
| 1340 |
+
"loss": 0.2187,
|
| 1341 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1342 |
+
"rewards/chosen": -2.046875,
|
| 1343 |
+
"rewards/margins": 2.15625,
|
| 1344 |
+
"rewards/rejected": -4.1875,
|
| 1345 |
+
"step": 870
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 1.0551558752997603,
|
| 1349 |
+
"grad_norm": 14.450058470084105,
|
| 1350 |
+
"learning_rate": 3.6028431808085294e-07,
|
| 1351 |
+
"logits/chosen": -2.25,
|
| 1352 |
+
"logits/rejected": -2.265625,
|
| 1353 |
+
"logps/chosen": -238.0,
|
| 1354 |
+
"logps/rejected": -258.0,
|
| 1355 |
+
"loss": 0.2253,
|
| 1356 |
+
"rewards/accuracies": 0.9375,
|
| 1357 |
+
"rewards/chosen": -1.984375,
|
| 1358 |
+
"rewards/margins": 1.953125,
|
| 1359 |
+
"rewards/rejected": -3.953125,
|
| 1360 |
+
"step": 880
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.0671462829736211,
|
| 1364 |
+
"grad_norm": 14.392869335197297,
|
| 1365 |
+
"learning_rate": 3.5806308307418926e-07,
|
| 1366 |
+
"logits/chosen": -2.375,
|
| 1367 |
+
"logits/rejected": -2.390625,
|
| 1368 |
+
"logps/chosen": -236.0,
|
| 1369 |
+
"logps/rejected": -262.0,
|
| 1370 |
+
"loss": 0.1993,
|
| 1371 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1372 |
+
"rewards/chosen": -2.15625,
|
| 1373 |
+
"rewards/margins": 2.25,
|
| 1374 |
+
"rewards/rejected": -4.40625,
|
| 1375 |
+
"step": 890
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.079136690647482,
|
| 1379 |
+
"grad_norm": 11.360043739301632,
|
| 1380 |
+
"learning_rate": 3.5584184806752554e-07,
|
| 1381 |
+
"logits/chosen": -2.3125,
|
| 1382 |
+
"logits/rejected": -2.3125,
|
| 1383 |
+
"logps/chosen": -224.0,
|
| 1384 |
+
"logps/rejected": -252.0,
|
| 1385 |
+
"loss": 0.2037,
|
| 1386 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1387 |
+
"rewards/chosen": -1.8984375,
|
| 1388 |
+
"rewards/margins": 2.390625,
|
| 1389 |
+
"rewards/rejected": -4.28125,
|
| 1390 |
+
"step": 900
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 1.091127098321343,
|
| 1394 |
+
"grad_norm": 17.249700366406504,
|
| 1395 |
+
"learning_rate": 3.536206130608618e-07,
|
| 1396 |
+
"logits/chosen": -2.296875,
|
| 1397 |
+
"logits/rejected": -2.328125,
|
| 1398 |
+
"logps/chosen": -224.0,
|
| 1399 |
+
"logps/rejected": -244.0,
|
| 1400 |
+
"loss": 0.1878,
|
| 1401 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1402 |
+
"rewards/chosen": -1.921875,
|
| 1403 |
+
"rewards/margins": 2.34375,
|
| 1404 |
+
"rewards/rejected": -4.28125,
|
| 1405 |
+
"step": 910
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 1.1031175059952039,
|
| 1409 |
+
"grad_norm": 17.500202516480442,
|
| 1410 |
+
"learning_rate": 3.513993780541981e-07,
|
| 1411 |
+
"logits/chosen": -2.28125,
|
| 1412 |
+
"logits/rejected": -2.375,
|
| 1413 |
+
"logps/chosen": -203.0,
|
| 1414 |
+
"logps/rejected": -246.0,
|
| 1415 |
+
"loss": 0.2128,
|
| 1416 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1417 |
+
"rewards/chosen": -1.875,
|
| 1418 |
+
"rewards/margins": 2.375,
|
| 1419 |
+
"rewards/rejected": -4.25,
|
| 1420 |
+
"step": 920
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 1.1151079136690647,
|
| 1424 |
+
"grad_norm": 15.16644938505708,
|
| 1425 |
+
"learning_rate": 3.491781430475344e-07,
|
| 1426 |
+
"logits/chosen": -2.359375,
|
| 1427 |
+
"logits/rejected": -2.40625,
|
| 1428 |
+
"logps/chosen": -220.0,
|
| 1429 |
+
"logps/rejected": -266.0,
|
| 1430 |
+
"loss": 0.1957,
|
| 1431 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1432 |
+
"rewards/chosen": -1.4375,
|
| 1433 |
+
"rewards/margins": 2.40625,
|
| 1434 |
+
"rewards/rejected": -3.84375,
|
| 1435 |
+
"step": 930
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 1.1270983213429258,
|
| 1439 |
+
"grad_norm": 12.741517880568063,
|
| 1440 |
+
"learning_rate": 3.469569080408707e-07,
|
| 1441 |
+
"logits/chosen": -2.21875,
|
| 1442 |
+
"logits/rejected": -2.234375,
|
| 1443 |
+
"logps/chosen": -225.0,
|
| 1444 |
+
"logps/rejected": -246.0,
|
| 1445 |
+
"loss": 0.2034,
|
| 1446 |
+
"rewards/accuracies": 0.9375,
|
| 1447 |
+
"rewards/chosen": -2.296875,
|
| 1448 |
+
"rewards/margins": 2.359375,
|
| 1449 |
+
"rewards/rejected": -4.65625,
|
| 1450 |
+
"step": 940
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 1.1390887290167866,
|
| 1454 |
+
"grad_norm": 21.470206899673656,
|
| 1455 |
+
"learning_rate": 3.4473567303420703e-07,
|
| 1456 |
+
"logits/chosen": -2.296875,
|
| 1457 |
+
"logits/rejected": -2.359375,
|
| 1458 |
+
"logps/chosen": -235.0,
|
| 1459 |
+
"logps/rejected": -256.0,
|
| 1460 |
+
"loss": 0.1846,
|
| 1461 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1462 |
+
"rewards/chosen": -2.34375,
|
| 1463 |
+
"rewards/margins": 2.515625,
|
| 1464 |
+
"rewards/rejected": -4.875,
|
| 1465 |
+
"step": 950
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.1510791366906474,
|
| 1469 |
+
"grad_norm": 17.20708415573527,
|
| 1470 |
+
"learning_rate": 3.425144380275433e-07,
|
| 1471 |
+
"logits/chosen": -2.265625,
|
| 1472 |
+
"logits/rejected": -2.359375,
|
| 1473 |
+
"logps/chosen": -245.0,
|
| 1474 |
+
"logps/rejected": -276.0,
|
| 1475 |
+
"loss": 0.2031,
|
| 1476 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1477 |
+
"rewards/chosen": -2.3125,
|
| 1478 |
+
"rewards/margins": 2.5625,
|
| 1479 |
+
"rewards/rejected": -4.875,
|
| 1480 |
+
"step": 960
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.1630695443645085,
|
| 1484 |
+
"grad_norm": 14.942195535135212,
|
| 1485 |
+
"learning_rate": 3.402932030208796e-07,
|
| 1486 |
+
"logits/chosen": -2.328125,
|
| 1487 |
+
"logits/rejected": -2.3125,
|
| 1488 |
+
"logps/chosen": -242.0,
|
| 1489 |
+
"logps/rejected": -272.0,
|
| 1490 |
+
"loss": 0.2032,
|
| 1491 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1492 |
+
"rewards/chosen": -2.234375,
|
| 1493 |
+
"rewards/margins": 2.734375,
|
| 1494 |
+
"rewards/rejected": -4.96875,
|
| 1495 |
+
"step": 970
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 1.1750599520383693,
|
| 1499 |
+
"grad_norm": 11.998202232926158,
|
| 1500 |
+
"learning_rate": 3.380719680142159e-07,
|
| 1501 |
+
"logits/chosen": -2.25,
|
| 1502 |
+
"logits/rejected": -2.328125,
|
| 1503 |
+
"logps/chosen": -226.0,
|
| 1504 |
+
"logps/rejected": -248.0,
|
| 1505 |
+
"loss": 0.1639,
|
| 1506 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1507 |
+
"rewards/chosen": -1.734375,
|
| 1508 |
+
"rewards/margins": 2.6875,
|
| 1509 |
+
"rewards/rejected": -4.40625,
|
| 1510 |
+
"step": 980
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 1.1870503597122302,
|
| 1514 |
+
"grad_norm": 12.340216646822865,
|
| 1515 |
+
"learning_rate": 3.358507330075522e-07,
|
| 1516 |
+
"logits/chosen": -2.25,
|
| 1517 |
+
"logits/rejected": -2.28125,
|
| 1518 |
+
"logps/chosen": -238.0,
|
| 1519 |
+
"logps/rejected": -260.0,
|
| 1520 |
+
"loss": 0.1902,
|
| 1521 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 1522 |
+
"rewards/chosen": -2.75,
|
| 1523 |
+
"rewards/margins": 2.296875,
|
| 1524 |
+
"rewards/rejected": -5.0625,
|
| 1525 |
+
"step": 990
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 1.1990407673860912,
|
| 1529 |
+
"grad_norm": 10.240701799756478,
|
| 1530 |
+
"learning_rate": 3.3362949800088847e-07,
|
| 1531 |
+
"logits/chosen": -2.34375,
|
| 1532 |
+
"logits/rejected": -2.28125,
|
| 1533 |
+
"logps/chosen": -237.0,
|
| 1534 |
+
"logps/rejected": -260.0,
|
| 1535 |
+
"loss": 0.1614,
|
| 1536 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1537 |
+
"rewards/chosen": -1.859375,
|
| 1538 |
+
"rewards/margins": 2.65625,
|
| 1539 |
+
"rewards/rejected": -4.5,
|
| 1540 |
+
"step": 1000
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 1.211031175059952,
|
| 1544 |
+
"grad_norm": 16.753004063840088,
|
| 1545 |
+
"learning_rate": 3.3140826299422474e-07,
|
| 1546 |
+
"logits/chosen": -2.3125,
|
| 1547 |
+
"logits/rejected": -2.390625,
|
| 1548 |
+
"logps/chosen": -234.0,
|
| 1549 |
+
"logps/rejected": -264.0,
|
| 1550 |
+
"loss": 0.2139,
|
| 1551 |
+
"rewards/accuracies": 0.9375,
|
| 1552 |
+
"rewards/chosen": -2.15625,
|
| 1553 |
+
"rewards/margins": 2.5625,
|
| 1554 |
+
"rewards/rejected": -4.71875,
|
| 1555 |
+
"step": 1010
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 1.223021582733813,
|
| 1559 |
+
"grad_norm": 22.84859803236856,
|
| 1560 |
+
"learning_rate": 3.291870279875611e-07,
|
| 1561 |
+
"logits/chosen": -2.25,
|
| 1562 |
+
"logits/rejected": -2.25,
|
| 1563 |
+
"logps/chosen": -242.0,
|
| 1564 |
+
"logps/rejected": -270.0,
|
| 1565 |
+
"loss": 0.1757,
|
| 1566 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1567 |
+
"rewards/chosen": -2.359375,
|
| 1568 |
+
"rewards/margins": 2.578125,
|
| 1569 |
+
"rewards/rejected": -4.9375,
|
| 1570 |
+
"step": 1020
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.235011990407674,
|
| 1574 |
+
"grad_norm": 14.789129276628799,
|
| 1575 |
+
"learning_rate": 3.269657929808974e-07,
|
| 1576 |
+
"logits/chosen": -2.296875,
|
| 1577 |
+
"logits/rejected": -2.296875,
|
| 1578 |
+
"logps/chosen": -234.0,
|
| 1579 |
+
"logps/rejected": -262.0,
|
| 1580 |
+
"loss": 0.1719,
|
| 1581 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1582 |
+
"rewards/chosen": -2.078125,
|
| 1583 |
+
"rewards/margins": 2.515625,
|
| 1584 |
+
"rewards/rejected": -4.59375,
|
| 1585 |
+
"step": 1030
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.2470023980815348,
|
| 1589 |
+
"grad_norm": 21.148440474411384,
|
| 1590 |
+
"learning_rate": 3.247445579742337e-07,
|
| 1591 |
+
"logits/chosen": -2.265625,
|
| 1592 |
+
"logits/rejected": -2.28125,
|
| 1593 |
+
"logps/chosen": -228.0,
|
| 1594 |
+
"logps/rejected": -266.0,
|
| 1595 |
+
"loss": 0.1689,
|
| 1596 |
+
"rewards/accuracies": 0.9375,
|
| 1597 |
+
"rewards/chosen": -2.6875,
|
| 1598 |
+
"rewards/margins": 2.546875,
|
| 1599 |
+
"rewards/rejected": -5.25,
|
| 1600 |
+
"step": 1040
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 1.2589928057553956,
|
| 1604 |
+
"grad_norm": 19.272034515617282,
|
| 1605 |
+
"learning_rate": 3.2252332296756996e-07,
|
| 1606 |
+
"logits/chosen": -2.28125,
|
| 1607 |
+
"logits/rejected": -2.328125,
|
| 1608 |
+
"logps/chosen": -227.0,
|
| 1609 |
+
"logps/rejected": -260.0,
|
| 1610 |
+
"loss": 0.1591,
|
| 1611 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1612 |
+
"rewards/chosen": -2.65625,
|
| 1613 |
+
"rewards/margins": 2.96875,
|
| 1614 |
+
"rewards/rejected": -5.625,
|
| 1615 |
+
"step": 1050
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 1.2709832134292567,
|
| 1619 |
+
"grad_norm": 16.85198782317795,
|
| 1620 |
+
"learning_rate": 3.2030208796090623e-07,
|
| 1621 |
+
"logits/chosen": -2.28125,
|
| 1622 |
+
"logits/rejected": -2.3125,
|
| 1623 |
+
"logps/chosen": -229.0,
|
| 1624 |
+
"logps/rejected": -251.0,
|
| 1625 |
+
"loss": 0.2236,
|
| 1626 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1627 |
+
"rewards/chosen": -1.765625,
|
| 1628 |
+
"rewards/margins": 2.890625,
|
| 1629 |
+
"rewards/rejected": -4.65625,
|
| 1630 |
+
"step": 1060
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 1.2829736211031175,
|
| 1634 |
+
"grad_norm": 15.836692596712956,
|
| 1635 |
+
"learning_rate": 3.180808529542425e-07,
|
| 1636 |
+
"logits/chosen": -2.25,
|
| 1637 |
+
"logits/rejected": -2.28125,
|
| 1638 |
+
"logps/chosen": -254.0,
|
| 1639 |
+
"logps/rejected": -276.0,
|
| 1640 |
+
"loss": 0.168,
|
| 1641 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1642 |
+
"rewards/chosen": -2.671875,
|
| 1643 |
+
"rewards/margins": 2.703125,
|
| 1644 |
+
"rewards/rejected": -5.375,
|
| 1645 |
+
"step": 1070
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 1.2949640287769784,
|
| 1649 |
+
"grad_norm": 8.91604867948452,
|
| 1650 |
+
"learning_rate": 3.1585961794757884e-07,
|
| 1651 |
+
"logits/chosen": -2.25,
|
| 1652 |
+
"logits/rejected": -2.28125,
|
| 1653 |
+
"logps/chosen": -240.0,
|
| 1654 |
+
"logps/rejected": -253.0,
|
| 1655 |
+
"loss": 0.1711,
|
| 1656 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1657 |
+
"rewards/chosen": -1.59375,
|
| 1658 |
+
"rewards/margins": 2.75,
|
| 1659 |
+
"rewards/rejected": -4.34375,
|
| 1660 |
+
"step": 1080
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 1.3069544364508392,
|
| 1664 |
+
"grad_norm": 14.305189630485472,
|
| 1665 |
+
"learning_rate": 3.1363838294091517e-07,
|
| 1666 |
+
"logits/chosen": -2.25,
|
| 1667 |
+
"logits/rejected": -2.28125,
|
| 1668 |
+
"logps/chosen": -235.0,
|
| 1669 |
+
"logps/rejected": -251.0,
|
| 1670 |
+
"loss": 0.1415,
|
| 1671 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1672 |
+
"rewards/chosen": -1.640625,
|
| 1673 |
+
"rewards/margins": 2.9375,
|
| 1674 |
+
"rewards/rejected": -4.59375,
|
| 1675 |
+
"step": 1090
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.3189448441247003,
|
| 1679 |
+
"grad_norm": 12.352735099035222,
|
| 1680 |
+
"learning_rate": 3.1141714793425145e-07,
|
| 1681 |
+
"logits/chosen": -2.296875,
|
| 1682 |
+
"logits/rejected": -2.28125,
|
| 1683 |
+
"logps/chosen": -249.0,
|
| 1684 |
+
"logps/rejected": -278.0,
|
| 1685 |
+
"loss": 0.1572,
|
| 1686 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1687 |
+
"rewards/chosen": -2.1875,
|
| 1688 |
+
"rewards/margins": 2.953125,
|
| 1689 |
+
"rewards/rejected": -5.125,
|
| 1690 |
+
"step": 1100
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.330935251798561,
|
| 1694 |
+
"grad_norm": 14.924171575427302,
|
| 1695 |
+
"learning_rate": 3.091959129275877e-07,
|
| 1696 |
+
"logits/chosen": -2.265625,
|
| 1697 |
+
"logits/rejected": -2.265625,
|
| 1698 |
+
"logps/chosen": -215.0,
|
| 1699 |
+
"logps/rejected": -241.0,
|
| 1700 |
+
"loss": 0.1714,
|
| 1701 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1702 |
+
"rewards/chosen": -1.9375,
|
| 1703 |
+
"rewards/margins": 2.84375,
|
| 1704 |
+
"rewards/rejected": -4.78125,
|
| 1705 |
+
"step": 1110
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 1.3429256594724222,
|
| 1709 |
+
"grad_norm": 17.499923046869892,
|
| 1710 |
+
"learning_rate": 3.06974677920924e-07,
|
| 1711 |
+
"logits/chosen": -2.296875,
|
| 1712 |
+
"logits/rejected": -2.296875,
|
| 1713 |
+
"logps/chosen": -254.0,
|
| 1714 |
+
"logps/rejected": -278.0,
|
| 1715 |
+
"loss": 0.1393,
|
| 1716 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1717 |
+
"rewards/chosen": -2.953125,
|
| 1718 |
+
"rewards/margins": 2.640625,
|
| 1719 |
+
"rewards/rejected": -5.59375,
|
| 1720 |
+
"step": 1120
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 1.354916067146283,
|
| 1724 |
+
"grad_norm": 16.158253954777205,
|
| 1725 |
+
"learning_rate": 3.0475344291426033e-07,
|
| 1726 |
+
"logits/chosen": -2.265625,
|
| 1727 |
+
"logits/rejected": -2.3125,
|
| 1728 |
+
"logps/chosen": -229.0,
|
| 1729 |
+
"logps/rejected": -266.0,
|
| 1730 |
+
"loss": 0.1656,
|
| 1731 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1732 |
+
"rewards/chosen": -2.1875,
|
| 1733 |
+
"rewards/margins": 3.375,
|
| 1734 |
+
"rewards/rejected": -5.5625,
|
| 1735 |
+
"step": 1130
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 1.3669064748201438,
|
| 1739 |
+
"grad_norm": 12.087846600129158,
|
| 1740 |
+
"learning_rate": 3.025322079075966e-07,
|
| 1741 |
+
"logits/chosen": -2.234375,
|
| 1742 |
+
"logits/rejected": -2.296875,
|
| 1743 |
+
"logps/chosen": -225.0,
|
| 1744 |
+
"logps/rejected": -272.0,
|
| 1745 |
+
"loss": 0.1966,
|
| 1746 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1747 |
+
"rewards/chosen": -1.96875,
|
| 1748 |
+
"rewards/margins": 3.171875,
|
| 1749 |
+
"rewards/rejected": -5.15625,
|
| 1750 |
+
"step": 1140
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 1.3788968824940047,
|
| 1754 |
+
"grad_norm": 17.576347022242885,
|
| 1755 |
+
"learning_rate": 3.003109729009329e-07,
|
| 1756 |
+
"logits/chosen": -2.25,
|
| 1757 |
+
"logits/rejected": -2.28125,
|
| 1758 |
+
"logps/chosen": -220.0,
|
| 1759 |
+
"logps/rejected": -264.0,
|
| 1760 |
+
"loss": 0.1847,
|
| 1761 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1762 |
+
"rewards/chosen": -1.8046875,
|
| 1763 |
+
"rewards/margins": 2.703125,
|
| 1764 |
+
"rewards/rejected": -4.5,
|
| 1765 |
+
"step": 1150
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 1.3908872901678657,
|
| 1769 |
+
"grad_norm": 12.233329087327826,
|
| 1770 |
+
"learning_rate": 2.980897378942692e-07,
|
| 1771 |
+
"logits/chosen": -2.28125,
|
| 1772 |
+
"logits/rejected": -2.234375,
|
| 1773 |
+
"logps/chosen": -235.0,
|
| 1774 |
+
"logps/rejected": -262.0,
|
| 1775 |
+
"loss": 0.1373,
|
| 1776 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1777 |
+
"rewards/chosen": -2.34375,
|
| 1778 |
+
"rewards/margins": 2.890625,
|
| 1779 |
+
"rewards/rejected": -5.21875,
|
| 1780 |
+
"step": 1160
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 1.4028776978417266,
|
| 1784 |
+
"grad_norm": 14.731501359525842,
|
| 1785 |
+
"learning_rate": 2.958685028876055e-07,
|
| 1786 |
+
"logits/chosen": -2.265625,
|
| 1787 |
+
"logits/rejected": -2.234375,
|
| 1788 |
+
"logps/chosen": -236.0,
|
| 1789 |
+
"logps/rejected": -260.0,
|
| 1790 |
+
"loss": 0.1843,
|
| 1791 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1792 |
+
"rewards/chosen": -2.1875,
|
| 1793 |
+
"rewards/margins": 2.625,
|
| 1794 |
+
"rewards/rejected": -4.8125,
|
| 1795 |
+
"step": 1170
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 1.4148681055155876,
|
| 1799 |
+
"grad_norm": 10.33569347198834,
|
| 1800 |
+
"learning_rate": 2.936472678809418e-07,
|
| 1801 |
+
"logits/chosen": -2.3125,
|
| 1802 |
+
"logits/rejected": -2.296875,
|
| 1803 |
+
"logps/chosen": -246.0,
|
| 1804 |
+
"logps/rejected": -266.0,
|
| 1805 |
+
"loss": 0.1912,
|
| 1806 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1807 |
+
"rewards/chosen": -2.40625,
|
| 1808 |
+
"rewards/margins": 2.640625,
|
| 1809 |
+
"rewards/rejected": -5.0625,
|
| 1810 |
+
"step": 1180
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 1.4268585131894485,
|
| 1814 |
+
"grad_norm": 16.986111059013577,
|
| 1815 |
+
"learning_rate": 2.914260328742781e-07,
|
| 1816 |
+
"logits/chosen": -2.25,
|
| 1817 |
+
"logits/rejected": -2.25,
|
| 1818 |
+
"logps/chosen": -231.0,
|
| 1819 |
+
"logps/rejected": -256.0,
|
| 1820 |
+
"loss": 0.1611,
|
| 1821 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1822 |
+
"rewards/chosen": -2.3125,
|
| 1823 |
+
"rewards/margins": 2.703125,
|
| 1824 |
+
"rewards/rejected": -5.03125,
|
| 1825 |
+
"step": 1190
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 1.4388489208633093,
|
| 1829 |
+
"grad_norm": 25.435636014827384,
|
| 1830 |
+
"learning_rate": 2.8920479786761437e-07,
|
| 1831 |
+
"logits/chosen": -2.28125,
|
| 1832 |
+
"logits/rejected": -2.296875,
|
| 1833 |
+
"logps/chosen": -235.0,
|
| 1834 |
+
"logps/rejected": -268.0,
|
| 1835 |
+
"loss": 0.1512,
|
| 1836 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1837 |
+
"rewards/chosen": -2.28125,
|
| 1838 |
+
"rewards/margins": 2.859375,
|
| 1839 |
+
"rewards/rejected": -5.15625,
|
| 1840 |
+
"step": 1200
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 1.4508393285371701,
|
| 1844 |
+
"grad_norm": 13.56680412334536,
|
| 1845 |
+
"learning_rate": 2.8698356286095065e-07,
|
| 1846 |
+
"logits/chosen": -2.234375,
|
| 1847 |
+
"logits/rejected": -2.265625,
|
| 1848 |
+
"logps/chosen": -260.0,
|
| 1849 |
+
"logps/rejected": -284.0,
|
| 1850 |
+
"loss": 0.1255,
|
| 1851 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1852 |
+
"rewards/chosen": -3.171875,
|
| 1853 |
+
"rewards/margins": 3.40625,
|
| 1854 |
+
"rewards/rejected": -6.59375,
|
| 1855 |
+
"step": 1210
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 1.4628297362110312,
|
| 1859 |
+
"grad_norm": 8.141641110335339,
|
| 1860 |
+
"learning_rate": 2.847623278542869e-07,
|
| 1861 |
+
"logits/chosen": -2.25,
|
| 1862 |
+
"logits/rejected": -2.3125,
|
| 1863 |
+
"logps/chosen": -236.0,
|
| 1864 |
+
"logps/rejected": -264.0,
|
| 1865 |
+
"loss": 0.1428,
|
| 1866 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1867 |
+
"rewards/chosen": -2.296875,
|
| 1868 |
+
"rewards/margins": 3.03125,
|
| 1869 |
+
"rewards/rejected": -5.34375,
|
| 1870 |
+
"step": 1220
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 1.474820143884892,
|
| 1874 |
+
"grad_norm": 21.115162166818017,
|
| 1875 |
+
"learning_rate": 2.825410928476233e-07,
|
| 1876 |
+
"logits/chosen": -2.1875,
|
| 1877 |
+
"logits/rejected": -2.25,
|
| 1878 |
+
"logps/chosen": -221.0,
|
| 1879 |
+
"logps/rejected": -268.0,
|
| 1880 |
+
"loss": 0.136,
|
| 1881 |
+
"rewards/accuracies": 0.9375,
|
| 1882 |
+
"rewards/chosen": -2.796875,
|
| 1883 |
+
"rewards/margins": 3.09375,
|
| 1884 |
+
"rewards/rejected": -5.90625,
|
| 1885 |
+
"step": 1230
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 1.486810551558753,
|
| 1889 |
+
"grad_norm": 18.47262881897785,
|
| 1890 |
+
"learning_rate": 2.803198578409596e-07,
|
| 1891 |
+
"logits/chosen": -2.25,
|
| 1892 |
+
"logits/rejected": -2.25,
|
| 1893 |
+
"logps/chosen": -228.0,
|
| 1894 |
+
"logps/rejected": -260.0,
|
| 1895 |
+
"loss": 0.1688,
|
| 1896 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1897 |
+
"rewards/chosen": -1.9375,
|
| 1898 |
+
"rewards/margins": 3.171875,
|
| 1899 |
+
"rewards/rejected": -5.125,
|
| 1900 |
+
"step": 1240
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 1.498800959232614,
|
| 1904 |
+
"grad_norm": 7.863109211524176,
|
| 1905 |
+
"learning_rate": 2.7809862283429586e-07,
|
| 1906 |
+
"logits/chosen": -2.234375,
|
| 1907 |
+
"logits/rejected": -2.296875,
|
| 1908 |
+
"logps/chosen": -222.0,
|
| 1909 |
+
"logps/rejected": -272.0,
|
| 1910 |
+
"loss": 0.1298,
|
| 1911 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1912 |
+
"rewards/chosen": -2.046875,
|
| 1913 |
+
"rewards/margins": 3.53125,
|
| 1914 |
+
"rewards/rejected": -5.5625,
|
| 1915 |
+
"step": 1250
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 1.5107913669064748,
|
| 1919 |
+
"grad_norm": 12.978026445303877,
|
| 1920 |
+
"learning_rate": 2.7587738782763214e-07,
|
| 1921 |
+
"logits/chosen": -2.265625,
|
| 1922 |
+
"logits/rejected": -2.328125,
|
| 1923 |
+
"logps/chosen": -216.0,
|
| 1924 |
+
"logps/rejected": -266.0,
|
| 1925 |
+
"loss": 0.1403,
|
| 1926 |
+
"rewards/accuracies": 0.9375,
|
| 1927 |
+
"rewards/chosen": -2.40625,
|
| 1928 |
+
"rewards/margins": 3.421875,
|
| 1929 |
+
"rewards/rejected": -5.8125,
|
| 1930 |
+
"step": 1260
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 1.5227817745803356,
|
| 1934 |
+
"grad_norm": 15.921496272567383,
|
| 1935 |
+
"learning_rate": 2.736561528209684e-07,
|
| 1936 |
+
"logits/chosen": -2.28125,
|
| 1937 |
+
"logits/rejected": -2.296875,
|
| 1938 |
+
"logps/chosen": -234.0,
|
| 1939 |
+
"logps/rejected": -274.0,
|
| 1940 |
+
"loss": 0.1493,
|
| 1941 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1942 |
+
"rewards/chosen": -2.28125,
|
| 1943 |
+
"rewards/margins": 3.140625,
|
| 1944 |
+
"rewards/rejected": -5.40625,
|
| 1945 |
+
"step": 1270
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 1.5347721822541966,
|
| 1949 |
+
"grad_norm": 14.590593841694169,
|
| 1950 |
+
"learning_rate": 2.7143491781430474e-07,
|
| 1951 |
+
"logits/chosen": -2.21875,
|
| 1952 |
+
"logits/rejected": -2.234375,
|
| 1953 |
+
"logps/chosen": -220.0,
|
| 1954 |
+
"logps/rejected": -268.0,
|
| 1955 |
+
"loss": 0.1147,
|
| 1956 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1957 |
+
"rewards/chosen": -2.8125,
|
| 1958 |
+
"rewards/margins": 3.078125,
|
| 1959 |
+
"rewards/rejected": -5.875,
|
| 1960 |
+
"step": 1280
|
| 1961 |
+
},
|
| 1962 |
+
{
|
| 1963 |
+
"epoch": 1.5467625899280577,
|
| 1964 |
+
"grad_norm": 16.15992031571578,
|
| 1965 |
+
"learning_rate": 2.692136828076411e-07,
|
| 1966 |
+
"logits/chosen": -2.296875,
|
| 1967 |
+
"logits/rejected": -2.28125,
|
| 1968 |
+
"logps/chosen": -242.0,
|
| 1969 |
+
"logps/rejected": -274.0,
|
| 1970 |
+
"loss": 0.1222,
|
| 1971 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1972 |
+
"rewards/chosen": -3.0625,
|
| 1973 |
+
"rewards/margins": 3.453125,
|
| 1974 |
+
"rewards/rejected": -6.53125,
|
| 1975 |
+
"step": 1290
|
| 1976 |
+
},
|
| 1977 |
+
{
|
| 1978 |
+
"epoch": 1.5587529976019185,
|
| 1979 |
+
"grad_norm": 10.104750544983435,
|
| 1980 |
+
"learning_rate": 2.6699244780097735e-07,
|
| 1981 |
+
"logits/chosen": -2.296875,
|
| 1982 |
+
"logits/rejected": -2.28125,
|
| 1983 |
+
"logps/chosen": -236.0,
|
| 1984 |
+
"logps/rejected": -266.0,
|
| 1985 |
+
"loss": 0.1225,
|
| 1986 |
+
"rewards/accuracies": 0.9375,
|
| 1987 |
+
"rewards/chosen": -2.453125,
|
| 1988 |
+
"rewards/margins": 3.34375,
|
| 1989 |
+
"rewards/rejected": -5.8125,
|
| 1990 |
+
"step": 1300
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 1.5707434052757794,
|
| 1994 |
+
"grad_norm": 16.414889130607445,
|
| 1995 |
+
"learning_rate": 2.6477121279431363e-07,
|
| 1996 |
+
"logits/chosen": -2.265625,
|
| 1997 |
+
"logits/rejected": -2.3125,
|
| 1998 |
+
"logps/chosen": -227.0,
|
| 1999 |
+
"logps/rejected": -270.0,
|
| 2000 |
+
"loss": 0.1865,
|
| 2001 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2002 |
+
"rewards/chosen": -2.59375,
|
| 2003 |
+
"rewards/margins": 3.421875,
|
| 2004 |
+
"rewards/rejected": -6.03125,
|
| 2005 |
+
"step": 1310
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 1.5827338129496402,
|
| 2009 |
+
"grad_norm": 19.612192416903497,
|
| 2010 |
+
"learning_rate": 2.625499777876499e-07,
|
| 2011 |
+
"logits/chosen": -2.265625,
|
| 2012 |
+
"logits/rejected": -2.28125,
|
| 2013 |
+
"logps/chosen": -236.0,
|
| 2014 |
+
"logps/rejected": -272.0,
|
| 2015 |
+
"loss": 0.1157,
|
| 2016 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2017 |
+
"rewards/chosen": -3.484375,
|
| 2018 |
+
"rewards/margins": 3.265625,
|
| 2019 |
+
"rewards/rejected": -6.75,
|
| 2020 |
+
"step": 1320
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 1.594724220623501,
|
| 2024 |
+
"grad_norm": 14.340619052654011,
|
| 2025 |
+
"learning_rate": 2.6032874278098623e-07,
|
| 2026 |
+
"logits/chosen": -2.3125,
|
| 2027 |
+
"logits/rejected": -2.296875,
|
| 2028 |
+
"logps/chosen": -251.0,
|
| 2029 |
+
"logps/rejected": -274.0,
|
| 2030 |
+
"loss": 0.1074,
|
| 2031 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2032 |
+
"rewards/chosen": -2.421875,
|
| 2033 |
+
"rewards/margins": 3.90625,
|
| 2034 |
+
"rewards/rejected": -6.34375,
|
| 2035 |
+
"step": 1330
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 1.6067146282973621,
|
| 2039 |
+
"grad_norm": 10.947155370834658,
|
| 2040 |
+
"learning_rate": 2.581075077743225e-07,
|
| 2041 |
+
"logits/chosen": -2.28125,
|
| 2042 |
+
"logits/rejected": -2.28125,
|
| 2043 |
+
"logps/chosen": -242.0,
|
| 2044 |
+
"logps/rejected": -268.0,
|
| 2045 |
+
"loss": 0.1142,
|
| 2046 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2047 |
+
"rewards/chosen": -2.53125,
|
| 2048 |
+
"rewards/margins": 3.640625,
|
| 2049 |
+
"rewards/rejected": -6.1875,
|
| 2050 |
+
"step": 1340
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 1.6187050359712232,
|
| 2054 |
+
"grad_norm": 6.379407877957059,
|
| 2055 |
+
"learning_rate": 2.558862727676588e-07,
|
| 2056 |
+
"logits/chosen": -2.3125,
|
| 2057 |
+
"logits/rejected": -2.296875,
|
| 2058 |
+
"logps/chosen": -243.0,
|
| 2059 |
+
"logps/rejected": -264.0,
|
| 2060 |
+
"loss": 0.1175,
|
| 2061 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2062 |
+
"rewards/chosen": -2.546875,
|
| 2063 |
+
"rewards/margins": 3.6875,
|
| 2064 |
+
"rewards/rejected": -6.21875,
|
| 2065 |
+
"step": 1350
|
| 2066 |
+
},
|
| 2067 |
+
{
|
| 2068 |
+
"epoch": 1.630695443645084,
|
| 2069 |
+
"grad_norm": 14.584427754669349,
|
| 2070 |
+
"learning_rate": 2.536650377609951e-07,
|
| 2071 |
+
"logits/chosen": -2.265625,
|
| 2072 |
+
"logits/rejected": -2.3125,
|
| 2073 |
+
"logps/chosen": -243.0,
|
| 2074 |
+
"logps/rejected": -268.0,
|
| 2075 |
+
"loss": 0.088,
|
| 2076 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2077 |
+
"rewards/chosen": -2.65625,
|
| 2078 |
+
"rewards/margins": 3.515625,
|
| 2079 |
+
"rewards/rejected": -6.1875,
|
| 2080 |
+
"step": 1360
|
| 2081 |
+
},
|
| 2082 |
+
{
|
| 2083 |
+
"epoch": 1.6426858513189448,
|
| 2084 |
+
"grad_norm": 18.386633865745754,
|
| 2085 |
+
"learning_rate": 2.514438027543314e-07,
|
| 2086 |
+
"logits/chosen": -2.234375,
|
| 2087 |
+
"logits/rejected": -2.203125,
|
| 2088 |
+
"logps/chosen": -240.0,
|
| 2089 |
+
"logps/rejected": -276.0,
|
| 2090 |
+
"loss": 0.0843,
|
| 2091 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2092 |
+
"rewards/chosen": -2.375,
|
| 2093 |
+
"rewards/margins": 4.1875,
|
| 2094 |
+
"rewards/rejected": -6.5625,
|
| 2095 |
+
"step": 1370
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 1.6546762589928057,
|
| 2099 |
+
"grad_norm": 13.249684798008296,
|
| 2100 |
+
"learning_rate": 2.492225677476677e-07,
|
| 2101 |
+
"logits/chosen": -2.296875,
|
| 2102 |
+
"logits/rejected": -2.328125,
|
| 2103 |
+
"logps/chosen": -230.0,
|
| 2104 |
+
"logps/rejected": -282.0,
|
| 2105 |
+
"loss": 0.11,
|
| 2106 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2107 |
+
"rewards/chosen": -2.515625,
|
| 2108 |
+
"rewards/margins": 3.46875,
|
| 2109 |
+
"rewards/rejected": -5.96875,
|
| 2110 |
+
"step": 1380
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 1.6666666666666665,
|
| 2114 |
+
"grad_norm": 24.989358834998974,
|
| 2115 |
+
"learning_rate": 2.47001332741004e-07,
|
| 2116 |
+
"logits/chosen": -2.265625,
|
| 2117 |
+
"logits/rejected": -2.3125,
|
| 2118 |
+
"logps/chosen": -240.0,
|
| 2119 |
+
"logps/rejected": -258.0,
|
| 2120 |
+
"loss": 0.134,
|
| 2121 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2122 |
+
"rewards/chosen": -2.59375,
|
| 2123 |
+
"rewards/margins": 3.3125,
|
| 2124 |
+
"rewards/rejected": -5.90625,
|
| 2125 |
+
"step": 1390
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 1.6786570743405276,
|
| 2129 |
+
"grad_norm": 21.225948587869443,
|
| 2130 |
+
"learning_rate": 2.447800977343403e-07,
|
| 2131 |
+
"logits/chosen": -2.203125,
|
| 2132 |
+
"logits/rejected": -2.21875,
|
| 2133 |
+
"logps/chosen": -227.0,
|
| 2134 |
+
"logps/rejected": -264.0,
|
| 2135 |
+
"loss": 0.1539,
|
| 2136 |
+
"rewards/accuracies": 0.9125000238418579,
|
| 2137 |
+
"rewards/chosen": -2.671875,
|
| 2138 |
+
"rewards/margins": 3.546875,
|
| 2139 |
+
"rewards/rejected": -6.21875,
|
| 2140 |
+
"step": 1400
|
| 2141 |
+
},
|
| 2142 |
+
{
|
| 2143 |
+
"epoch": 1.6906474820143886,
|
| 2144 |
+
"grad_norm": 22.09693803294425,
|
| 2145 |
+
"learning_rate": 2.425588627276766e-07,
|
| 2146 |
+
"logits/chosen": -2.265625,
|
| 2147 |
+
"logits/rejected": -2.28125,
|
| 2148 |
+
"logps/chosen": -232.0,
|
| 2149 |
+
"logps/rejected": -274.0,
|
| 2150 |
+
"loss": 0.1301,
|
| 2151 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2152 |
+
"rewards/chosen": -2.8125,
|
| 2153 |
+
"rewards/margins": 3.59375,
|
| 2154 |
+
"rewards/rejected": -6.40625,
|
| 2155 |
+
"step": 1410
|
| 2156 |
+
},
|
| 2157 |
+
{
|
| 2158 |
+
"epoch": 1.7026378896882495,
|
| 2159 |
+
"grad_norm": 13.568164124099978,
|
| 2160 |
+
"learning_rate": 2.403376277210129e-07,
|
| 2161 |
+
"logits/chosen": -2.25,
|
| 2162 |
+
"logits/rejected": -2.296875,
|
| 2163 |
+
"logps/chosen": -245.0,
|
| 2164 |
+
"logps/rejected": -286.0,
|
| 2165 |
+
"loss": 0.0997,
|
| 2166 |
+
"rewards/accuracies": 1.0,
|
| 2167 |
+
"rewards/chosen": -3.734375,
|
| 2168 |
+
"rewards/margins": 3.609375,
|
| 2169 |
+
"rewards/rejected": -7.34375,
|
| 2170 |
+
"step": 1420
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 1.7146282973621103,
|
| 2174 |
+
"grad_norm": 12.536257234153933,
|
| 2175 |
+
"learning_rate": 2.3811639271434916e-07,
|
| 2176 |
+
"logits/chosen": -2.265625,
|
| 2177 |
+
"logits/rejected": -2.25,
|
| 2178 |
+
"logps/chosen": -243.0,
|
| 2179 |
+
"logps/rejected": -274.0,
|
| 2180 |
+
"loss": 0.1365,
|
| 2181 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2182 |
+
"rewards/chosen": -2.765625,
|
| 2183 |
+
"rewards/margins": 3.515625,
|
| 2184 |
+
"rewards/rejected": -6.28125,
|
| 2185 |
+
"step": 1430
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"epoch": 1.7266187050359711,
|
| 2189 |
+
"grad_norm": 17.74910297920457,
|
| 2190 |
+
"learning_rate": 2.3589515770768546e-07,
|
| 2191 |
+
"logits/chosen": -2.296875,
|
| 2192 |
+
"logits/rejected": -2.296875,
|
| 2193 |
+
"logps/chosen": -256.0,
|
| 2194 |
+
"logps/rejected": -276.0,
|
| 2195 |
+
"loss": 0.1436,
|
| 2196 |
+
"rewards/accuracies": 0.8999999761581421,
|
| 2197 |
+
"rewards/chosen": -3.3125,
|
| 2198 |
+
"rewards/margins": 3.53125,
|
| 2199 |
+
"rewards/rejected": -6.84375,
|
| 2200 |
+
"step": 1440
|
| 2201 |
+
},
|
| 2202 |
+
{
|
| 2203 |
+
"epoch": 1.738609112709832,
|
| 2204 |
+
"grad_norm": 27.091464814087082,
|
| 2205 |
+
"learning_rate": 2.3367392270102177e-07,
|
| 2206 |
+
"logits/chosen": -2.25,
|
| 2207 |
+
"logits/rejected": -2.28125,
|
| 2208 |
+
"logps/chosen": -235.0,
|
| 2209 |
+
"logps/rejected": -298.0,
|
| 2210 |
+
"loss": 0.1031,
|
| 2211 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2212 |
+
"rewards/chosen": -2.640625,
|
| 2213 |
+
"rewards/margins": 4.1875,
|
| 2214 |
+
"rewards/rejected": -6.84375,
|
| 2215 |
+
"step": 1450
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 1.750599520383693,
|
| 2219 |
+
"grad_norm": 14.870002835228663,
|
| 2220 |
+
"learning_rate": 2.3145268769435804e-07,
|
| 2221 |
+
"logits/chosen": -2.265625,
|
| 2222 |
+
"logits/rejected": -2.203125,
|
| 2223 |
+
"logps/chosen": -247.0,
|
| 2224 |
+
"logps/rejected": -278.0,
|
| 2225 |
+
"loss": 0.1465,
|
| 2226 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2227 |
+
"rewards/chosen": -3.046875,
|
| 2228 |
+
"rewards/margins": 3.609375,
|
| 2229 |
+
"rewards/rejected": -6.65625,
|
| 2230 |
+
"step": 1460
|
| 2231 |
+
},
|
| 2232 |
+
{
|
| 2233 |
+
"epoch": 1.762589928057554,
|
| 2234 |
+
"grad_norm": 17.480506529993356,
|
| 2235 |
+
"learning_rate": 2.2923145268769435e-07,
|
| 2236 |
+
"logits/chosen": -2.25,
|
| 2237 |
+
"logits/rejected": -2.265625,
|
| 2238 |
+
"logps/chosen": -223.0,
|
| 2239 |
+
"logps/rejected": -264.0,
|
| 2240 |
+
"loss": 0.0896,
|
| 2241 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2242 |
+
"rewards/chosen": -2.484375,
|
| 2243 |
+
"rewards/margins": 4.3125,
|
| 2244 |
+
"rewards/rejected": -6.78125,
|
| 2245 |
+
"step": 1470
|
| 2246 |
+
},
|
| 2247 |
+
{
|
| 2248 |
+
"epoch": 1.774580335731415,
|
| 2249 |
+
"grad_norm": 6.540878635676034,
|
| 2250 |
+
"learning_rate": 2.2701021768103065e-07,
|
| 2251 |
+
"logits/chosen": -2.25,
|
| 2252 |
+
"logits/rejected": -2.25,
|
| 2253 |
+
"logps/chosen": -236.0,
|
| 2254 |
+
"logps/rejected": -278.0,
|
| 2255 |
+
"loss": 0.1107,
|
| 2256 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2257 |
+
"rewards/chosen": -2.703125,
|
| 2258 |
+
"rewards/margins": 3.984375,
|
| 2259 |
+
"rewards/rejected": -6.6875,
|
| 2260 |
+
"step": 1480
|
| 2261 |
+
},
|
| 2262 |
+
{
|
| 2263 |
+
"epoch": 1.7865707434052758,
|
| 2264 |
+
"grad_norm": 16.51964209338177,
|
| 2265 |
+
"learning_rate": 2.2478898267436695e-07,
|
| 2266 |
+
"logits/chosen": -2.28125,
|
| 2267 |
+
"logits/rejected": -2.265625,
|
| 2268 |
+
"logps/chosen": -241.0,
|
| 2269 |
+
"logps/rejected": -284.0,
|
| 2270 |
+
"loss": 0.0963,
|
| 2271 |
+
"rewards/accuracies": 1.0,
|
| 2272 |
+
"rewards/chosen": -3.671875,
|
| 2273 |
+
"rewards/margins": 3.71875,
|
| 2274 |
+
"rewards/rejected": -7.375,
|
| 2275 |
+
"step": 1490
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 1.7985611510791366,
|
| 2279 |
+
"grad_norm": 17.145067233621255,
|
| 2280 |
+
"learning_rate": 2.2256774766770323e-07,
|
| 2281 |
+
"logits/chosen": -2.1875,
|
| 2282 |
+
"logits/rejected": -2.171875,
|
| 2283 |
+
"logps/chosen": -234.0,
|
| 2284 |
+
"logps/rejected": -255.0,
|
| 2285 |
+
"loss": 0.1016,
|
| 2286 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2287 |
+
"rewards/chosen": -2.71875,
|
| 2288 |
+
"rewards/margins": 3.953125,
|
| 2289 |
+
"rewards/rejected": -6.6875,
|
| 2290 |
+
"step": 1500
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 1.8105515587529974,
|
| 2294 |
+
"grad_norm": 9.418365551008751,
|
| 2295 |
+
"learning_rate": 2.2034651266103953e-07,
|
| 2296 |
+
"logits/chosen": -2.234375,
|
| 2297 |
+
"logits/rejected": -2.234375,
|
| 2298 |
+
"logps/chosen": -216.0,
|
| 2299 |
+
"logps/rejected": -266.0,
|
| 2300 |
+
"loss": 0.0902,
|
| 2301 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2302 |
+
"rewards/chosen": -2.65625,
|
| 2303 |
+
"rewards/margins": 3.875,
|
| 2304 |
+
"rewards/rejected": -6.53125,
|
| 2305 |
+
"step": 1510
|
| 2306 |
+
},
|
| 2307 |
+
{
|
| 2308 |
+
"epoch": 1.8225419664268585,
|
| 2309 |
+
"grad_norm": 8.454143934854207,
|
| 2310 |
+
"learning_rate": 2.1812527765437583e-07,
|
| 2311 |
+
"logits/chosen": -2.265625,
|
| 2312 |
+
"logits/rejected": -2.25,
|
| 2313 |
+
"logps/chosen": -247.0,
|
| 2314 |
+
"logps/rejected": -278.0,
|
| 2315 |
+
"loss": 0.1036,
|
| 2316 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2317 |
+
"rewards/chosen": -2.875,
|
| 2318 |
+
"rewards/margins": 3.75,
|
| 2319 |
+
"rewards/rejected": -6.625,
|
| 2320 |
+
"step": 1520
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 1.8345323741007196,
|
| 2324 |
+
"grad_norm": 23.877830813468428,
|
| 2325 |
+
"learning_rate": 2.159040426477121e-07,
|
| 2326 |
+
"logits/chosen": -2.28125,
|
| 2327 |
+
"logits/rejected": -2.3125,
|
| 2328 |
+
"logps/chosen": -249.0,
|
| 2329 |
+
"logps/rejected": -300.0,
|
| 2330 |
+
"loss": 0.0974,
|
| 2331 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2332 |
+
"rewards/chosen": -3.625,
|
| 2333 |
+
"rewards/margins": 3.953125,
|
| 2334 |
+
"rewards/rejected": -7.5625,
|
| 2335 |
+
"step": 1530
|
| 2336 |
+
},
|
| 2337 |
+
{
|
| 2338 |
+
"epoch": 1.8465227817745804,
|
| 2339 |
+
"grad_norm": 17.59044264644789,
|
| 2340 |
+
"learning_rate": 2.1368280764104841e-07,
|
| 2341 |
+
"logits/chosen": -2.171875,
|
| 2342 |
+
"logits/rejected": -2.296875,
|
| 2343 |
+
"logps/chosen": -242.0,
|
| 2344 |
+
"logps/rejected": -300.0,
|
| 2345 |
+
"loss": 0.0921,
|
| 2346 |
+
"rewards/accuracies": 1.0,
|
| 2347 |
+
"rewards/chosen": -3.734375,
|
| 2348 |
+
"rewards/margins": 3.984375,
|
| 2349 |
+
"rewards/rejected": -7.71875,
|
| 2350 |
+
"step": 1540
|
| 2351 |
+
},
|
| 2352 |
+
{
|
| 2353 |
+
"epoch": 1.8585131894484412,
|
| 2354 |
+
"grad_norm": 7.53671069585816,
|
| 2355 |
+
"learning_rate": 2.1146157263438472e-07,
|
| 2356 |
+
"logits/chosen": -2.265625,
|
| 2357 |
+
"logits/rejected": -2.265625,
|
| 2358 |
+
"logps/chosen": -237.0,
|
| 2359 |
+
"logps/rejected": -278.0,
|
| 2360 |
+
"loss": 0.0677,
|
| 2361 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2362 |
+
"rewards/chosen": -2.734375,
|
| 2363 |
+
"rewards/margins": 4.1875,
|
| 2364 |
+
"rewards/rejected": -6.9375,
|
| 2365 |
+
"step": 1550
|
| 2366 |
+
},
|
| 2367 |
+
{
|
| 2368 |
+
"epoch": 1.870503597122302,
|
| 2369 |
+
"grad_norm": 15.114334538388968,
|
| 2370 |
+
"learning_rate": 2.09240337627721e-07,
|
| 2371 |
+
"logits/chosen": -2.28125,
|
| 2372 |
+
"logits/rejected": -2.25,
|
| 2373 |
+
"logps/chosen": -245.0,
|
| 2374 |
+
"logps/rejected": -260.0,
|
| 2375 |
+
"loss": 0.1064,
|
| 2376 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 2377 |
+
"rewards/chosen": -3.0,
|
| 2378 |
+
"rewards/margins": 3.78125,
|
| 2379 |
+
"rewards/rejected": -6.78125,
|
| 2380 |
+
"step": 1560
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 1.882494004796163,
|
| 2384 |
+
"grad_norm": 24.921093039405555,
|
| 2385 |
+
"learning_rate": 2.070191026210573e-07,
|
| 2386 |
+
"logits/chosen": -2.25,
|
| 2387 |
+
"logits/rejected": -2.296875,
|
| 2388 |
+
"logps/chosen": -246.0,
|
| 2389 |
+
"logps/rejected": -274.0,
|
| 2390 |
+
"loss": 0.1109,
|
| 2391 |
+
"rewards/accuracies": 0.949999988079071,
|
| 2392 |
+
"rewards/chosen": -2.6875,
|
| 2393 |
+
"rewards/margins": 3.46875,
|
| 2394 |
+
"rewards/rejected": -6.15625,
|
| 2395 |
+
"step": 1570
|
| 2396 |
+
},
|
| 2397 |
+
{
|
| 2398 |
+
"epoch": 1.894484412470024,
|
| 2399 |
+
"grad_norm": 17.229671646251834,
|
| 2400 |
+
"learning_rate": 2.047978676143936e-07,
|
| 2401 |
+
"logits/chosen": -2.203125,
|
| 2402 |
+
"logits/rejected": -2.21875,
|
| 2403 |
+
"logps/chosen": -251.0,
|
| 2404 |
+
"logps/rejected": -276.0,
|
| 2405 |
+
"loss": 0.0787,
|
| 2406 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2407 |
+
"rewards/chosen": -2.609375,
|
| 2408 |
+
"rewards/margins": 3.875,
|
| 2409 |
+
"rewards/rejected": -6.5,
|
| 2410 |
+
"step": 1580
|
| 2411 |
+
},
|
| 2412 |
+
{
|
| 2413 |
+
"epoch": 1.906474820143885,
|
| 2414 |
+
"grad_norm": 10.939803601324265,
|
| 2415 |
+
"learning_rate": 2.025766326077299e-07,
|
| 2416 |
+
"logits/chosen": -2.265625,
|
| 2417 |
+
"logits/rejected": -2.265625,
|
| 2418 |
+
"logps/chosen": -235.0,
|
| 2419 |
+
"logps/rejected": -284.0,
|
| 2420 |
+
"loss": 0.0871,
|
| 2421 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2422 |
+
"rewards/chosen": -2.515625,
|
| 2423 |
+
"rewards/margins": 4.125,
|
| 2424 |
+
"rewards/rejected": -6.625,
|
| 2425 |
+
"step": 1590
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 1.9184652278177459,
|
| 2429 |
+
"grad_norm": 16.05247518882454,
|
| 2430 |
+
"learning_rate": 2.0035539760106618e-07,
|
| 2431 |
+
"logits/chosen": -2.265625,
|
| 2432 |
+
"logits/rejected": -2.296875,
|
| 2433 |
+
"logps/chosen": -241.0,
|
| 2434 |
+
"logps/rejected": -300.0,
|
| 2435 |
+
"loss": 0.0611,
|
| 2436 |
+
"rewards/accuracies": 1.0,
|
| 2437 |
+
"rewards/chosen": -2.5625,
|
| 2438 |
+
"rewards/margins": 4.78125,
|
| 2439 |
+
"rewards/rejected": -7.34375,
|
| 2440 |
+
"step": 1600
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 1.9304556354916067,
|
| 2444 |
+
"grad_norm": 9.855796335839392,
|
| 2445 |
+
"learning_rate": 1.9813416259440246e-07,
|
| 2446 |
+
"logits/chosen": -2.234375,
|
| 2447 |
+
"logits/rejected": -2.28125,
|
| 2448 |
+
"logps/chosen": -245.0,
|
| 2449 |
+
"logps/rejected": -278.0,
|
| 2450 |
+
"loss": 0.0889,
|
| 2451 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2452 |
+
"rewards/chosen": -3.625,
|
| 2453 |
+
"rewards/margins": 3.921875,
|
| 2454 |
+
"rewards/rejected": -7.53125,
|
| 2455 |
+
"step": 1610
|
| 2456 |
+
},
|
| 2457 |
+
{
|
| 2458 |
+
"epoch": 1.9424460431654675,
|
| 2459 |
+
"grad_norm": 7.909249841156511,
|
| 2460 |
+
"learning_rate": 1.9591292758773879e-07,
|
| 2461 |
+
"logits/chosen": -2.25,
|
| 2462 |
+
"logits/rejected": -2.296875,
|
| 2463 |
+
"logps/chosen": -235.0,
|
| 2464 |
+
"logps/rejected": -298.0,
|
| 2465 |
+
"loss": 0.1069,
|
| 2466 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2467 |
+
"rewards/chosen": -3.59375,
|
| 2468 |
+
"rewards/margins": 4.40625,
|
| 2469 |
+
"rewards/rejected": -8.0,
|
| 2470 |
+
"step": 1620
|
| 2471 |
+
},
|
| 2472 |
+
{
|
| 2473 |
+
"epoch": 1.9544364508393284,
|
| 2474 |
+
"grad_norm": 10.195924724786721,
|
| 2475 |
+
"learning_rate": 1.9369169258107506e-07,
|
| 2476 |
+
"logits/chosen": -2.328125,
|
| 2477 |
+
"logits/rejected": -2.234375,
|
| 2478 |
+
"logps/chosen": -251.0,
|
| 2479 |
+
"logps/rejected": -280.0,
|
| 2480 |
+
"loss": 0.0671,
|
| 2481 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 2482 |
+
"rewards/chosen": -3.296875,
|
| 2483 |
+
"rewards/margins": 4.0625,
|
| 2484 |
+
"rewards/rejected": -7.375,
|
| 2485 |
+
"step": 1630
|
| 2486 |
+
},
|
| 2487 |
+
{
|
| 2488 |
+
"epoch": 1.9664268585131894,
|
| 2489 |
+
"grad_norm": 4.981347556467333,
|
| 2490 |
+
"learning_rate": 1.9147045757441137e-07,
|
| 2491 |
+
"logits/chosen": -2.265625,
|
| 2492 |
+
"logits/rejected": -2.296875,
|
| 2493 |
+
"logps/chosen": -232.0,
|
| 2494 |
+
"logps/rejected": -286.0,
|
| 2495 |
+
"loss": 0.0626,
|
| 2496 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2497 |
+
"rewards/chosen": -3.40625,
|
| 2498 |
+
"rewards/margins": 4.03125,
|
| 2499 |
+
"rewards/rejected": -7.4375,
|
| 2500 |
+
"step": 1640
|
| 2501 |
+
},
|
| 2502 |
+
{
|
| 2503 |
+
"epoch": 1.9784172661870505,
|
| 2504 |
+
"grad_norm": 8.780872639436975,
|
| 2505 |
+
"learning_rate": 1.8924922256774767e-07,
|
| 2506 |
+
"logits/chosen": -2.25,
|
| 2507 |
+
"logits/rejected": -2.1875,
|
| 2508 |
+
"logps/chosen": -248.0,
|
| 2509 |
+
"logps/rejected": -276.0,
|
| 2510 |
+
"loss": 0.0836,
|
| 2511 |
+
"rewards/accuracies": 0.987500011920929,
|
| 2512 |
+
"rewards/chosen": -3.84375,
|
| 2513 |
+
"rewards/margins": 3.953125,
|
| 2514 |
+
"rewards/rejected": -7.8125,
|
| 2515 |
+
"step": 1650
|
| 2516 |
+
},
|
| 2517 |
+
{
|
| 2518 |
+
"epoch": 1.9904076738609113,
|
| 2519 |
+
"grad_norm": 13.586444001383855,
|
| 2520 |
+
"learning_rate": 1.8702798756108395e-07,
|
| 2521 |
+
"logits/chosen": -2.34375,
|
| 2522 |
+
"logits/rejected": -2.265625,
|
| 2523 |
+
"logps/chosen": -253.0,
|
| 2524 |
+
"logps/rejected": -282.0,
|
| 2525 |
+
"loss": 0.0676,
|
| 2526 |
+
"rewards/accuracies": 1.0,
|
| 2527 |
+
"rewards/chosen": -3.046875,
|
| 2528 |
+
"rewards/margins": 4.21875,
|
| 2529 |
+
"rewards/rejected": -7.25,
|
| 2530 |
+
"step": 1660
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 2.0,
|
| 2534 |
+
"eval_logits/chosen": -2.265625,
|
| 2535 |
+
"eval_logits/rejected": -2.296875,
|
| 2536 |
+
"eval_logps/chosen": -255.0,
|
| 2537 |
+
"eval_logps/rejected": -270.0,
|
| 2538 |
+
"eval_loss": 0.8723406195640564,
|
| 2539 |
+
"eval_rewards/accuracies": 0.6397058963775635,
|
| 2540 |
+
"eval_rewards/chosen": -4.71875,
|
| 2541 |
+
"eval_rewards/margins": 1.59375,
|
| 2542 |
+
"eval_rewards/rejected": -6.3125,
|
| 2543 |
+
"eval_runtime": 20.0713,
|
| 2544 |
+
"eval_samples_per_second": 19.979,
|
| 2545 |
+
"eval_steps_per_second": 0.847,
|
| 2546 |
+
"step": 1668
|
| 2547 |
}
|
| 2548 |
],
|
| 2549 |
"logging_steps": 10,
|