Training in progress, epoch 3, checkpoint
Browse files- last-checkpoint/global_step3744/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step3744/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step3744/zero_pp_rank_0_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step3744/zero_pp_rank_1_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/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +1894 -3
last-checkpoint/global_step3744/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 sha256:10aa22d0b86834e271c2260d25e9d0e6a2220dae19d9787828e4348b1421e2f7
|
| 3 |
+
size 30462473157
|
last-checkpoint/global_step3744/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 sha256:2b1dd7b8f42c293ec88f22157ab743fda1e47b9342f250d6719a9f6f39b8d778
|
| 3 |
+
size 30462473157
|
last-checkpoint/global_step3744/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 sha256:7ac3896df02945578505e5be29a4b6b7c9a2e582aaed23c9ae20b3db17eab617
|
| 3 |
+
size 168021
|
last-checkpoint/global_step3744/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:a000715579fa90fb876d81d3a862e534315f1dcd52c014cd773aa4327cab5466
|
| 3 |
+
size 168021
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step3744
|
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 sha256:5177a8e6782bb88bda2135ebe4b556b112bb813fd6ff6a0be384c71f8e0189d6
|
| 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:fc8950fe3195f952346a0b52c6dfe6413f42e8dbbef2ad1916bd02c754b27342
|
| 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 sha256:6fb232eefc76f6ffc2aa58fe324177383ef7b1c3560473ced7dfbf1ccf253d47
|
| 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:cabea12f73243814a74bfd1b213f64fb4c6348bad93b08cd5641d428fd19614e
|
| 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 sha256:95430508d31bbe1a66a940e2572bf04addefae3e1c4e861e8657f66d302aa23e
|
| 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:b70a6983e6085768b4d2b447a8bd35374bce9cf4ea8e8fbefc1260ca2e054a70
|
| 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:aa515306b1c7f0b288d157b662096d041c14c3bbec98c937898173d64b5dfffa
|
| 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,
|
|
@@ -3789,6 +3789,1897 @@
|
|
| 3789 |
"eval_samples_per_second": 7.686,
|
| 3790 |
"eval_steps_per_second": 0.966,
|
| 3791 |
"step": 2496
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3792 |
}
|
| 3793 |
],
|
| 3794 |
"logging_steps": 10,
|
|
@@ -3803,7 +5694,7 @@
|
|
| 3803 |
"should_evaluate": false,
|
| 3804 |
"should_log": false,
|
| 3805 |
"should_save": true,
|
| 3806 |
-
"should_training_stop":
|
| 3807 |
},
|
| 3808 |
"attributes": {}
|
| 3809 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 3744,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 3789 |
"eval_samples_per_second": 7.686,
|
| 3790 |
"eval_steps_per_second": 0.966,
|
| 3791 |
"step": 2496
|
| 3792 |
+
},
|
| 3793 |
+
{
|
| 3794 |
+
"epoch": 2.003205128205128,
|
| 3795 |
+
"grad_norm": 4.521723680059775e-05,
|
| 3796 |
+
"learning_rate": 1.84624517661027e-07,
|
| 3797 |
+
"logits/chosen": 0.8828125,
|
| 3798 |
+
"logits/rejected": 1.203125,
|
| 3799 |
+
"logps/chosen": -166.0,
|
| 3800 |
+
"logps/rejected": -320.0,
|
| 3801 |
+
"loss": 0.0,
|
| 3802 |
+
"rewards/accuracies": 1.0,
|
| 3803 |
+
"rewards/chosen": 3.625,
|
| 3804 |
+
"rewards/margins": 22.875,
|
| 3805 |
+
"rewards/rejected": -19.25,
|
| 3806 |
+
"step": 2500
|
| 3807 |
+
},
|
| 3808 |
+
{
|
| 3809 |
+
"epoch": 2.011217948717949,
|
| 3810 |
+
"grad_norm": 3.030267966817534e-06,
|
| 3811 |
+
"learning_rate": 1.8314039774413774e-07,
|
| 3812 |
+
"logits/chosen": 0.640625,
|
| 3813 |
+
"logits/rejected": 1.1171875,
|
| 3814 |
+
"logps/chosen": -195.0,
|
| 3815 |
+
"logps/rejected": -332.0,
|
| 3816 |
+
"loss": 0.0,
|
| 3817 |
+
"rewards/accuracies": 1.0,
|
| 3818 |
+
"rewards/chosen": 3.28125,
|
| 3819 |
+
"rewards/margins": 23.125,
|
| 3820 |
+
"rewards/rejected": -19.875,
|
| 3821 |
+
"step": 2510
|
| 3822 |
+
},
|
| 3823 |
+
{
|
| 3824 |
+
"epoch": 2.019230769230769,
|
| 3825 |
+
"grad_norm": 6.205803823998353e-07,
|
| 3826 |
+
"learning_rate": 1.8165627782724842e-07,
|
| 3827 |
+
"logits/chosen": 0.796875,
|
| 3828 |
+
"logits/rejected": 1.1328125,
|
| 3829 |
+
"logps/chosen": -141.0,
|
| 3830 |
+
"logps/rejected": -340.0,
|
| 3831 |
+
"loss": 0.0,
|
| 3832 |
+
"rewards/accuracies": 1.0,
|
| 3833 |
+
"rewards/chosen": 3.3125,
|
| 3834 |
+
"rewards/margins": 23.875,
|
| 3835 |
+
"rewards/rejected": -20.5,
|
| 3836 |
+
"step": 2520
|
| 3837 |
+
},
|
| 3838 |
+
{
|
| 3839 |
+
"epoch": 2.02724358974359,
|
| 3840 |
+
"grad_norm": 1.5474883916264918e-05,
|
| 3841 |
+
"learning_rate": 1.8017215791035914e-07,
|
| 3842 |
+
"logits/chosen": 0.87109375,
|
| 3843 |
+
"logits/rejected": 1.375,
|
| 3844 |
+
"logps/chosen": -172.0,
|
| 3845 |
+
"logps/rejected": -338.0,
|
| 3846 |
+
"loss": 0.0,
|
| 3847 |
+
"rewards/accuracies": 1.0,
|
| 3848 |
+
"rewards/chosen": 3.21875,
|
| 3849 |
+
"rewards/margins": 23.25,
|
| 3850 |
+
"rewards/rejected": -20.125,
|
| 3851 |
+
"step": 2530
|
| 3852 |
+
},
|
| 3853 |
+
{
|
| 3854 |
+
"epoch": 2.03525641025641,
|
| 3855 |
+
"grad_norm": 3.136476138411595e-06,
|
| 3856 |
+
"learning_rate": 1.7868803799346988e-07,
|
| 3857 |
+
"logits/chosen": 0.9140625,
|
| 3858 |
+
"logits/rejected": 1.2578125,
|
| 3859 |
+
"logps/chosen": -159.0,
|
| 3860 |
+
"logps/rejected": -326.0,
|
| 3861 |
+
"loss": 0.0,
|
| 3862 |
+
"rewards/accuracies": 1.0,
|
| 3863 |
+
"rewards/chosen": 3.890625,
|
| 3864 |
+
"rewards/margins": 23.0,
|
| 3865 |
+
"rewards/rejected": -19.125,
|
| 3866 |
+
"step": 2540
|
| 3867 |
+
},
|
| 3868 |
+
{
|
| 3869 |
+
"epoch": 2.043269230769231,
|
| 3870 |
+
"grad_norm": 2.0185773664202382e-07,
|
| 3871 |
+
"learning_rate": 1.772039180765806e-07,
|
| 3872 |
+
"logits/chosen": 0.8203125,
|
| 3873 |
+
"logits/rejected": 1.46875,
|
| 3874 |
+
"logps/chosen": -157.0,
|
| 3875 |
+
"logps/rejected": -316.0,
|
| 3876 |
+
"loss": 0.0,
|
| 3877 |
+
"rewards/accuracies": 1.0,
|
| 3878 |
+
"rewards/chosen": 3.96875,
|
| 3879 |
+
"rewards/margins": 22.0,
|
| 3880 |
+
"rewards/rejected": -18.0,
|
| 3881 |
+
"step": 2550
|
| 3882 |
+
},
|
| 3883 |
+
{
|
| 3884 |
+
"epoch": 2.051282051282051,
|
| 3885 |
+
"grad_norm": 6.762579805129387e-06,
|
| 3886 |
+
"learning_rate": 1.757197981596913e-07,
|
| 3887 |
+
"logits/chosen": 0.6953125,
|
| 3888 |
+
"logits/rejected": 1.2578125,
|
| 3889 |
+
"logps/chosen": -175.0,
|
| 3890 |
+
"logps/rejected": -320.0,
|
| 3891 |
+
"loss": 0.0,
|
| 3892 |
+
"rewards/accuracies": 1.0,
|
| 3893 |
+
"rewards/chosen": 3.828125,
|
| 3894 |
+
"rewards/margins": 22.375,
|
| 3895 |
+
"rewards/rejected": -18.5,
|
| 3896 |
+
"step": 2560
|
| 3897 |
+
},
|
| 3898 |
+
{
|
| 3899 |
+
"epoch": 2.059294871794872,
|
| 3900 |
+
"grad_norm": 1.272394096308946e-06,
|
| 3901 |
+
"learning_rate": 1.74235678242802e-07,
|
| 3902 |
+
"logits/chosen": 0.9921875,
|
| 3903 |
+
"logits/rejected": 1.2734375,
|
| 3904 |
+
"logps/chosen": -168.0,
|
| 3905 |
+
"logps/rejected": -338.0,
|
| 3906 |
+
"loss": 0.0,
|
| 3907 |
+
"rewards/accuracies": 1.0,
|
| 3908 |
+
"rewards/chosen": 3.6875,
|
| 3909 |
+
"rewards/margins": 23.125,
|
| 3910 |
+
"rewards/rejected": -19.5,
|
| 3911 |
+
"step": 2570
|
| 3912 |
+
},
|
| 3913 |
+
{
|
| 3914 |
+
"epoch": 2.0673076923076925,
|
| 3915 |
+
"grad_norm": 0.00022224801984980412,
|
| 3916 |
+
"learning_rate": 1.7275155832591273e-07,
|
| 3917 |
+
"logits/chosen": 0.86328125,
|
| 3918 |
+
"logits/rejected": 1.2734375,
|
| 3919 |
+
"logps/chosen": -124.0,
|
| 3920 |
+
"logps/rejected": -326.0,
|
| 3921 |
+
"loss": 0.0,
|
| 3922 |
+
"rewards/accuracies": 1.0,
|
| 3923 |
+
"rewards/chosen": 3.484375,
|
| 3924 |
+
"rewards/margins": 23.5,
|
| 3925 |
+
"rewards/rejected": -20.0,
|
| 3926 |
+
"step": 2580
|
| 3927 |
+
},
|
| 3928 |
+
{
|
| 3929 |
+
"epoch": 2.0753205128205128,
|
| 3930 |
+
"grad_norm": 5.872515304133134e-06,
|
| 3931 |
+
"learning_rate": 1.7126743840902344e-07,
|
| 3932 |
+
"logits/chosen": 0.9140625,
|
| 3933 |
+
"logits/rejected": 1.15625,
|
| 3934 |
+
"logps/chosen": -160.0,
|
| 3935 |
+
"logps/rejected": -318.0,
|
| 3936 |
+
"loss": 0.0001,
|
| 3937 |
+
"rewards/accuracies": 1.0,
|
| 3938 |
+
"rewards/chosen": 3.734375,
|
| 3939 |
+
"rewards/margins": 22.5,
|
| 3940 |
+
"rewards/rejected": -18.875,
|
| 3941 |
+
"step": 2590
|
| 3942 |
+
},
|
| 3943 |
+
{
|
| 3944 |
+
"epoch": 2.0833333333333335,
|
| 3945 |
+
"grad_norm": 0.0015141429378734963,
|
| 3946 |
+
"learning_rate": 1.6978331849213416e-07,
|
| 3947 |
+
"logits/chosen": 0.7578125,
|
| 3948 |
+
"logits/rejected": 1.1484375,
|
| 3949 |
+
"logps/chosen": -183.0,
|
| 3950 |
+
"logps/rejected": -336.0,
|
| 3951 |
+
"loss": 0.0,
|
| 3952 |
+
"rewards/accuracies": 1.0,
|
| 3953 |
+
"rewards/chosen": 3.328125,
|
| 3954 |
+
"rewards/margins": 23.125,
|
| 3955 |
+
"rewards/rejected": -19.875,
|
| 3956 |
+
"step": 2600
|
| 3957 |
+
},
|
| 3958 |
+
{
|
| 3959 |
+
"epoch": 2.0913461538461537,
|
| 3960 |
+
"grad_norm": 2.991210518482803e-06,
|
| 3961 |
+
"learning_rate": 1.6829919857524487e-07,
|
| 3962 |
+
"logits/chosen": 0.81640625,
|
| 3963 |
+
"logits/rejected": 0.8671875,
|
| 3964 |
+
"logps/chosen": -140.0,
|
| 3965 |
+
"logps/rejected": -342.0,
|
| 3966 |
+
"loss": 0.0,
|
| 3967 |
+
"rewards/accuracies": 1.0,
|
| 3968 |
+
"rewards/chosen": 2.96875,
|
| 3969 |
+
"rewards/margins": 23.5,
|
| 3970 |
+
"rewards/rejected": -20.5,
|
| 3971 |
+
"step": 2610
|
| 3972 |
+
},
|
| 3973 |
+
{
|
| 3974 |
+
"epoch": 2.0993589743589745,
|
| 3975 |
+
"grad_norm": 5.742496354978091e-06,
|
| 3976 |
+
"learning_rate": 1.6681507865835558e-07,
|
| 3977 |
+
"logits/chosen": 0.6953125,
|
| 3978 |
+
"logits/rejected": 1.3203125,
|
| 3979 |
+
"logps/chosen": -186.0,
|
| 3980 |
+
"logps/rejected": -338.0,
|
| 3981 |
+
"loss": 0.0,
|
| 3982 |
+
"rewards/accuracies": 1.0,
|
| 3983 |
+
"rewards/chosen": 3.265625,
|
| 3984 |
+
"rewards/margins": 23.125,
|
| 3985 |
+
"rewards/rejected": -19.875,
|
| 3986 |
+
"step": 2620
|
| 3987 |
+
},
|
| 3988 |
+
{
|
| 3989 |
+
"epoch": 2.1073717948717947,
|
| 3990 |
+
"grad_norm": 3.757183821571958e-05,
|
| 3991 |
+
"learning_rate": 1.653309587414663e-07,
|
| 3992 |
+
"logits/chosen": 0.98046875,
|
| 3993 |
+
"logits/rejected": 1.3046875,
|
| 3994 |
+
"logps/chosen": -150.0,
|
| 3995 |
+
"logps/rejected": -316.0,
|
| 3996 |
+
"loss": 0.0,
|
| 3997 |
+
"rewards/accuracies": 1.0,
|
| 3998 |
+
"rewards/chosen": 3.46875,
|
| 3999 |
+
"rewards/margins": 21.875,
|
| 4000 |
+
"rewards/rejected": -18.375,
|
| 4001 |
+
"step": 2630
|
| 4002 |
+
},
|
| 4003 |
+
{
|
| 4004 |
+
"epoch": 2.1153846153846154,
|
| 4005 |
+
"grad_norm": 1.1309302661788921e-05,
|
| 4006 |
+
"learning_rate": 1.63846838824577e-07,
|
| 4007 |
+
"logits/chosen": 0.796875,
|
| 4008 |
+
"logits/rejected": 1.21875,
|
| 4009 |
+
"logps/chosen": -195.0,
|
| 4010 |
+
"logps/rejected": -330.0,
|
| 4011 |
+
"loss": 0.0,
|
| 4012 |
+
"rewards/accuracies": 1.0,
|
| 4013 |
+
"rewards/chosen": 3.015625,
|
| 4014 |
+
"rewards/margins": 22.125,
|
| 4015 |
+
"rewards/rejected": -19.125,
|
| 4016 |
+
"step": 2640
|
| 4017 |
+
},
|
| 4018 |
+
{
|
| 4019 |
+
"epoch": 2.123397435897436,
|
| 4020 |
+
"grad_norm": 0.00011609396836253147,
|
| 4021 |
+
"learning_rate": 1.6236271890768775e-07,
|
| 4022 |
+
"logits/chosen": 0.8828125,
|
| 4023 |
+
"logits/rejected": 1.578125,
|
| 4024 |
+
"logps/chosen": -143.0,
|
| 4025 |
+
"logps/rejected": -334.0,
|
| 4026 |
+
"loss": 0.0,
|
| 4027 |
+
"rewards/accuracies": 1.0,
|
| 4028 |
+
"rewards/chosen": 3.53125,
|
| 4029 |
+
"rewards/margins": 23.125,
|
| 4030 |
+
"rewards/rejected": -19.625,
|
| 4031 |
+
"step": 2650
|
| 4032 |
+
},
|
| 4033 |
+
{
|
| 4034 |
+
"epoch": 2.1314102564102564,
|
| 4035 |
+
"grad_norm": 2.37606345324713e-07,
|
| 4036 |
+
"learning_rate": 1.6087859899079844e-07,
|
| 4037 |
+
"logits/chosen": 0.703125,
|
| 4038 |
+
"logits/rejected": 1.40625,
|
| 4039 |
+
"logps/chosen": -175.0,
|
| 4040 |
+
"logps/rejected": -334.0,
|
| 4041 |
+
"loss": 0.0,
|
| 4042 |
+
"rewards/accuracies": 1.0,
|
| 4043 |
+
"rewards/chosen": 3.71875,
|
| 4044 |
+
"rewards/margins": 24.0,
|
| 4045 |
+
"rewards/rejected": -20.25,
|
| 4046 |
+
"step": 2660
|
| 4047 |
+
},
|
| 4048 |
+
{
|
| 4049 |
+
"epoch": 2.139423076923077,
|
| 4050 |
+
"grad_norm": 1.7790769529114564e-05,
|
| 4051 |
+
"learning_rate": 1.5939447907390915e-07,
|
| 4052 |
+
"logits/chosen": 0.59765625,
|
| 4053 |
+
"logits/rejected": 1.21875,
|
| 4054 |
+
"logps/chosen": -207.0,
|
| 4055 |
+
"logps/rejected": -352.0,
|
| 4056 |
+
"loss": 0.0,
|
| 4057 |
+
"rewards/accuracies": 1.0,
|
| 4058 |
+
"rewards/chosen": 3.453125,
|
| 4059 |
+
"rewards/margins": 23.625,
|
| 4060 |
+
"rewards/rejected": -20.25,
|
| 4061 |
+
"step": 2670
|
| 4062 |
+
},
|
| 4063 |
+
{
|
| 4064 |
+
"epoch": 2.1474358974358974,
|
| 4065 |
+
"grad_norm": 1.2324246884581975e-07,
|
| 4066 |
+
"learning_rate": 1.579103591570199e-07,
|
| 4067 |
+
"logits/chosen": 0.9296875,
|
| 4068 |
+
"logits/rejected": 1.3828125,
|
| 4069 |
+
"logps/chosen": -185.0,
|
| 4070 |
+
"logps/rejected": -334.0,
|
| 4071 |
+
"loss": 0.0,
|
| 4072 |
+
"rewards/accuracies": 1.0,
|
| 4073 |
+
"rewards/chosen": 3.34375,
|
| 4074 |
+
"rewards/margins": 23.125,
|
| 4075 |
+
"rewards/rejected": -19.75,
|
| 4076 |
+
"step": 2680
|
| 4077 |
+
},
|
| 4078 |
+
{
|
| 4079 |
+
"epoch": 2.155448717948718,
|
| 4080 |
+
"grad_norm": 8.415390201667563e-05,
|
| 4081 |
+
"learning_rate": 1.564262392401306e-07,
|
| 4082 |
+
"logits/chosen": 0.828125,
|
| 4083 |
+
"logits/rejected": 1.421875,
|
| 4084 |
+
"logps/chosen": -173.0,
|
| 4085 |
+
"logps/rejected": -338.0,
|
| 4086 |
+
"loss": 0.0,
|
| 4087 |
+
"rewards/accuracies": 1.0,
|
| 4088 |
+
"rewards/chosen": 3.5625,
|
| 4089 |
+
"rewards/margins": 24.125,
|
| 4090 |
+
"rewards/rejected": -20.5,
|
| 4091 |
+
"step": 2690
|
| 4092 |
+
},
|
| 4093 |
+
{
|
| 4094 |
+
"epoch": 2.1634615384615383,
|
| 4095 |
+
"grad_norm": 0.0011339204404605804,
|
| 4096 |
+
"learning_rate": 1.549421193232413e-07,
|
| 4097 |
+
"logits/chosen": 0.82421875,
|
| 4098 |
+
"logits/rejected": 1.15625,
|
| 4099 |
+
"logps/chosen": -168.0,
|
| 4100 |
+
"logps/rejected": -342.0,
|
| 4101 |
+
"loss": 0.0,
|
| 4102 |
+
"rewards/accuracies": 1.0,
|
| 4103 |
+
"rewards/chosen": 3.578125,
|
| 4104 |
+
"rewards/margins": 23.875,
|
| 4105 |
+
"rewards/rejected": -20.25,
|
| 4106 |
+
"step": 2700
|
| 4107 |
+
},
|
| 4108 |
+
{
|
| 4109 |
+
"epoch": 2.171474358974359,
|
| 4110 |
+
"grad_norm": 6.851096339481569e-06,
|
| 4111 |
+
"learning_rate": 1.5345799940635203e-07,
|
| 4112 |
+
"logits/chosen": 0.53125,
|
| 4113 |
+
"logits/rejected": 1.2734375,
|
| 4114 |
+
"logps/chosen": -197.0,
|
| 4115 |
+
"logps/rejected": -348.0,
|
| 4116 |
+
"loss": 0.0,
|
| 4117 |
+
"rewards/accuracies": 1.0,
|
| 4118 |
+
"rewards/chosen": 3.328125,
|
| 4119 |
+
"rewards/margins": 24.25,
|
| 4120 |
+
"rewards/rejected": -20.875,
|
| 4121 |
+
"step": 2710
|
| 4122 |
+
},
|
| 4123 |
+
{
|
| 4124 |
+
"epoch": 2.1794871794871793,
|
| 4125 |
+
"grad_norm": 0.0003726374461810356,
|
| 4126 |
+
"learning_rate": 1.5197387948946275e-07,
|
| 4127 |
+
"logits/chosen": 1.03125,
|
| 4128 |
+
"logits/rejected": 1.453125,
|
| 4129 |
+
"logps/chosen": -169.0,
|
| 4130 |
+
"logps/rejected": -320.0,
|
| 4131 |
+
"loss": 0.0,
|
| 4132 |
+
"rewards/accuracies": 1.0,
|
| 4133 |
+
"rewards/chosen": 3.8125,
|
| 4134 |
+
"rewards/margins": 22.5,
|
| 4135 |
+
"rewards/rejected": -18.75,
|
| 4136 |
+
"step": 2720
|
| 4137 |
+
},
|
| 4138 |
+
{
|
| 4139 |
+
"epoch": 2.1875,
|
| 4140 |
+
"grad_norm": 2.8940339945841385e-05,
|
| 4141 |
+
"learning_rate": 1.5048975957257346e-07,
|
| 4142 |
+
"logits/chosen": 0.78125,
|
| 4143 |
+
"logits/rejected": 1.1171875,
|
| 4144 |
+
"logps/chosen": -168.0,
|
| 4145 |
+
"logps/rejected": -340.0,
|
| 4146 |
+
"loss": 0.0,
|
| 4147 |
+
"rewards/accuracies": 1.0,
|
| 4148 |
+
"rewards/chosen": 2.96875,
|
| 4149 |
+
"rewards/margins": 23.125,
|
| 4150 |
+
"rewards/rejected": -20.125,
|
| 4151 |
+
"step": 2730
|
| 4152 |
+
},
|
| 4153 |
+
{
|
| 4154 |
+
"epoch": 2.1955128205128207,
|
| 4155 |
+
"grad_norm": 0.002859780208342844,
|
| 4156 |
+
"learning_rate": 1.4900563965568417e-07,
|
| 4157 |
+
"logits/chosen": 0.8671875,
|
| 4158 |
+
"logits/rejected": 1.4140625,
|
| 4159 |
+
"logps/chosen": -171.0,
|
| 4160 |
+
"logps/rejected": -312.0,
|
| 4161 |
+
"loss": 0.0,
|
| 4162 |
+
"rewards/accuracies": 1.0,
|
| 4163 |
+
"rewards/chosen": 3.453125,
|
| 4164 |
+
"rewards/margins": 21.0,
|
| 4165 |
+
"rewards/rejected": -17.625,
|
| 4166 |
+
"step": 2740
|
| 4167 |
+
},
|
| 4168 |
+
{
|
| 4169 |
+
"epoch": 2.203525641025641,
|
| 4170 |
+
"grad_norm": 8.642451938502949e-05,
|
| 4171 |
+
"learning_rate": 1.4752151973879489e-07,
|
| 4172 |
+
"logits/chosen": 0.92578125,
|
| 4173 |
+
"logits/rejected": 1.0703125,
|
| 4174 |
+
"logps/chosen": -152.0,
|
| 4175 |
+
"logps/rejected": -322.0,
|
| 4176 |
+
"loss": 0.0,
|
| 4177 |
+
"rewards/accuracies": 1.0,
|
| 4178 |
+
"rewards/chosen": 3.625,
|
| 4179 |
+
"rewards/margins": 23.125,
|
| 4180 |
+
"rewards/rejected": -19.5,
|
| 4181 |
+
"step": 2750
|
| 4182 |
+
},
|
| 4183 |
+
{
|
| 4184 |
+
"epoch": 2.2115384615384617,
|
| 4185 |
+
"grad_norm": 0.004973797903281045,
|
| 4186 |
+
"learning_rate": 1.460373998219056e-07,
|
| 4187 |
+
"logits/chosen": 1.171875,
|
| 4188 |
+
"logits/rejected": 1.5859375,
|
| 4189 |
+
"logps/chosen": -154.0,
|
| 4190 |
+
"logps/rejected": -312.0,
|
| 4191 |
+
"loss": 0.0,
|
| 4192 |
+
"rewards/accuracies": 1.0,
|
| 4193 |
+
"rewards/chosen": 3.296875,
|
| 4194 |
+
"rewards/margins": 22.625,
|
| 4195 |
+
"rewards/rejected": -19.25,
|
| 4196 |
+
"step": 2760
|
| 4197 |
+
},
|
| 4198 |
+
{
|
| 4199 |
+
"epoch": 2.219551282051282,
|
| 4200 |
+
"grad_norm": 0.0007041585520470051,
|
| 4201 |
+
"learning_rate": 1.4455327990501631e-07,
|
| 4202 |
+
"logits/chosen": 0.796875,
|
| 4203 |
+
"logits/rejected": 1.015625,
|
| 4204 |
+
"logps/chosen": -140.0,
|
| 4205 |
+
"logps/rejected": -338.0,
|
| 4206 |
+
"loss": 0.0,
|
| 4207 |
+
"rewards/accuracies": 1.0,
|
| 4208 |
+
"rewards/chosen": 3.421875,
|
| 4209 |
+
"rewards/margins": 24.75,
|
| 4210 |
+
"rewards/rejected": -21.25,
|
| 4211 |
+
"step": 2770
|
| 4212 |
+
},
|
| 4213 |
+
{
|
| 4214 |
+
"epoch": 2.2275641025641026,
|
| 4215 |
+
"grad_norm": 0.00045000223905260884,
|
| 4216 |
+
"learning_rate": 1.4306915998812705e-07,
|
| 4217 |
+
"logits/chosen": 1.0625,
|
| 4218 |
+
"logits/rejected": 1.4375,
|
| 4219 |
+
"logps/chosen": -171.0,
|
| 4220 |
+
"logps/rejected": -336.0,
|
| 4221 |
+
"loss": 0.0,
|
| 4222 |
+
"rewards/accuracies": 1.0,
|
| 4223 |
+
"rewards/chosen": 3.328125,
|
| 4224 |
+
"rewards/margins": 23.375,
|
| 4225 |
+
"rewards/rejected": -20.0,
|
| 4226 |
+
"step": 2780
|
| 4227 |
+
},
|
| 4228 |
+
{
|
| 4229 |
+
"epoch": 2.235576923076923,
|
| 4230 |
+
"grad_norm": 3.7708664164834145e-07,
|
| 4231 |
+
"learning_rate": 1.4158504007123777e-07,
|
| 4232 |
+
"logits/chosen": 0.921875,
|
| 4233 |
+
"logits/rejected": 1.34375,
|
| 4234 |
+
"logps/chosen": -188.0,
|
| 4235 |
+
"logps/rejected": -308.0,
|
| 4236 |
+
"loss": 0.0,
|
| 4237 |
+
"rewards/accuracies": 1.0,
|
| 4238 |
+
"rewards/chosen": 3.359375,
|
| 4239 |
+
"rewards/margins": 22.0,
|
| 4240 |
+
"rewards/rejected": -18.75,
|
| 4241 |
+
"step": 2790
|
| 4242 |
+
},
|
| 4243 |
+
{
|
| 4244 |
+
"epoch": 2.2435897435897436,
|
| 4245 |
+
"grad_norm": 2.820076649705414e-06,
|
| 4246 |
+
"learning_rate": 1.4010092015434845e-07,
|
| 4247 |
+
"logits/chosen": 0.765625,
|
| 4248 |
+
"logits/rejected": 1.375,
|
| 4249 |
+
"logps/chosen": -211.0,
|
| 4250 |
+
"logps/rejected": -330.0,
|
| 4251 |
+
"loss": 0.0,
|
| 4252 |
+
"rewards/accuracies": 1.0,
|
| 4253 |
+
"rewards/chosen": 3.703125,
|
| 4254 |
+
"rewards/margins": 22.625,
|
| 4255 |
+
"rewards/rejected": -19.0,
|
| 4256 |
+
"step": 2800
|
| 4257 |
+
},
|
| 4258 |
+
{
|
| 4259 |
+
"epoch": 2.251602564102564,
|
| 4260 |
+
"grad_norm": 7.476346576947421e-06,
|
| 4261 |
+
"learning_rate": 1.3861680023745917e-07,
|
| 4262 |
+
"logits/chosen": 0.96875,
|
| 4263 |
+
"logits/rejected": 1.171875,
|
| 4264 |
+
"logps/chosen": -150.0,
|
| 4265 |
+
"logps/rejected": -342.0,
|
| 4266 |
+
"loss": 0.0,
|
| 4267 |
+
"rewards/accuracies": 1.0,
|
| 4268 |
+
"rewards/chosen": 2.984375,
|
| 4269 |
+
"rewards/margins": 23.5,
|
| 4270 |
+
"rewards/rejected": -20.5,
|
| 4271 |
+
"step": 2810
|
| 4272 |
+
},
|
| 4273 |
+
{
|
| 4274 |
+
"epoch": 2.2596153846153846,
|
| 4275 |
+
"grad_norm": 7.318508636096753e-07,
|
| 4276 |
+
"learning_rate": 1.371326803205699e-07,
|
| 4277 |
+
"logits/chosen": 0.94140625,
|
| 4278 |
+
"logits/rejected": 1.4296875,
|
| 4279 |
+
"logps/chosen": -164.0,
|
| 4280 |
+
"logps/rejected": -322.0,
|
| 4281 |
+
"loss": 0.0,
|
| 4282 |
+
"rewards/accuracies": 1.0,
|
| 4283 |
+
"rewards/chosen": 3.46875,
|
| 4284 |
+
"rewards/margins": 22.625,
|
| 4285 |
+
"rewards/rejected": -19.125,
|
| 4286 |
+
"step": 2820
|
| 4287 |
+
},
|
| 4288 |
+
{
|
| 4289 |
+
"epoch": 2.2676282051282053,
|
| 4290 |
+
"grad_norm": 4.8320214702453866e-05,
|
| 4291 |
+
"learning_rate": 1.3564856040368062e-07,
|
| 4292 |
+
"logits/chosen": 0.94140625,
|
| 4293 |
+
"logits/rejected": 1.21875,
|
| 4294 |
+
"logps/chosen": -160.0,
|
| 4295 |
+
"logps/rejected": -332.0,
|
| 4296 |
+
"loss": 0.0,
|
| 4297 |
+
"rewards/accuracies": 1.0,
|
| 4298 |
+
"rewards/chosen": 3.5,
|
| 4299 |
+
"rewards/margins": 22.75,
|
| 4300 |
+
"rewards/rejected": -19.25,
|
| 4301 |
+
"step": 2830
|
| 4302 |
+
},
|
| 4303 |
+
{
|
| 4304 |
+
"epoch": 2.2756410256410255,
|
| 4305 |
+
"grad_norm": 5.030183317968072e-06,
|
| 4306 |
+
"learning_rate": 1.341644404867913e-07,
|
| 4307 |
+
"logits/chosen": 0.76953125,
|
| 4308 |
+
"logits/rejected": 1.0703125,
|
| 4309 |
+
"logps/chosen": -185.0,
|
| 4310 |
+
"logps/rejected": -328.0,
|
| 4311 |
+
"loss": 0.0005,
|
| 4312 |
+
"rewards/accuracies": 1.0,
|
| 4313 |
+
"rewards/chosen": 3.671875,
|
| 4314 |
+
"rewards/margins": 23.0,
|
| 4315 |
+
"rewards/rejected": -19.25,
|
| 4316 |
+
"step": 2840
|
| 4317 |
+
},
|
| 4318 |
+
{
|
| 4319 |
+
"epoch": 2.2836538461538463,
|
| 4320 |
+
"grad_norm": 0.00014825087080426772,
|
| 4321 |
+
"learning_rate": 1.3268032056990205e-07,
|
| 4322 |
+
"logits/chosen": 0.78515625,
|
| 4323 |
+
"logits/rejected": 1.3203125,
|
| 4324 |
+
"logps/chosen": -181.0,
|
| 4325 |
+
"logps/rejected": -338.0,
|
| 4326 |
+
"loss": 0.0,
|
| 4327 |
+
"rewards/accuracies": 1.0,
|
| 4328 |
+
"rewards/chosen": 3.796875,
|
| 4329 |
+
"rewards/margins": 23.5,
|
| 4330 |
+
"rewards/rejected": -19.75,
|
| 4331 |
+
"step": 2850
|
| 4332 |
+
},
|
| 4333 |
+
{
|
| 4334 |
+
"epoch": 2.2916666666666665,
|
| 4335 |
+
"grad_norm": 1.3011868588303958e-06,
|
| 4336 |
+
"learning_rate": 1.3119620065301276e-07,
|
| 4337 |
+
"logits/chosen": 1.0625,
|
| 4338 |
+
"logits/rejected": 1.3359375,
|
| 4339 |
+
"logps/chosen": -146.0,
|
| 4340 |
+
"logps/rejected": -332.0,
|
| 4341 |
+
"loss": 0.0,
|
| 4342 |
+
"rewards/accuracies": 1.0,
|
| 4343 |
+
"rewards/chosen": 3.09375,
|
| 4344 |
+
"rewards/margins": 23.25,
|
| 4345 |
+
"rewards/rejected": -20.125,
|
| 4346 |
+
"step": 2860
|
| 4347 |
+
},
|
| 4348 |
+
{
|
| 4349 |
+
"epoch": 2.2996794871794872,
|
| 4350 |
+
"grad_norm": 4.246488981116065e-06,
|
| 4351 |
+
"learning_rate": 1.2971208073612347e-07,
|
| 4352 |
+
"logits/chosen": 0.86328125,
|
| 4353 |
+
"logits/rejected": 1.4921875,
|
| 4354 |
+
"logps/chosen": -184.0,
|
| 4355 |
+
"logps/rejected": -332.0,
|
| 4356 |
+
"loss": 0.0,
|
| 4357 |
+
"rewards/accuracies": 1.0,
|
| 4358 |
+
"rewards/chosen": 3.578125,
|
| 4359 |
+
"rewards/margins": 22.875,
|
| 4360 |
+
"rewards/rejected": -19.25,
|
| 4361 |
+
"step": 2870
|
| 4362 |
+
},
|
| 4363 |
+
{
|
| 4364 |
+
"epoch": 2.3076923076923075,
|
| 4365 |
+
"grad_norm": 7.270042186092499e-06,
|
| 4366 |
+
"learning_rate": 1.282279608192342e-07,
|
| 4367 |
+
"logits/chosen": 0.66015625,
|
| 4368 |
+
"logits/rejected": 1.390625,
|
| 4369 |
+
"logps/chosen": -144.0,
|
| 4370 |
+
"logps/rejected": -326.0,
|
| 4371 |
+
"loss": 0.0,
|
| 4372 |
+
"rewards/accuracies": 1.0,
|
| 4373 |
+
"rewards/chosen": 3.203125,
|
| 4374 |
+
"rewards/margins": 23.75,
|
| 4375 |
+
"rewards/rejected": -20.5,
|
| 4376 |
+
"step": 2880
|
| 4377 |
+
},
|
| 4378 |
+
{
|
| 4379 |
+
"epoch": 2.315705128205128,
|
| 4380 |
+
"grad_norm": 1.4457369601091267e-08,
|
| 4381 |
+
"learning_rate": 1.267438409023449e-07,
|
| 4382 |
+
"logits/chosen": 1.0703125,
|
| 4383 |
+
"logits/rejected": 1.296875,
|
| 4384 |
+
"logps/chosen": -122.5,
|
| 4385 |
+
"logps/rejected": -336.0,
|
| 4386 |
+
"loss": 0.0,
|
| 4387 |
+
"rewards/accuracies": 1.0,
|
| 4388 |
+
"rewards/chosen": 3.015625,
|
| 4389 |
+
"rewards/margins": 23.375,
|
| 4390 |
+
"rewards/rejected": -20.25,
|
| 4391 |
+
"step": 2890
|
| 4392 |
+
},
|
| 4393 |
+
{
|
| 4394 |
+
"epoch": 2.323717948717949,
|
| 4395 |
+
"grad_norm": 1.2540641729400349e-06,
|
| 4396 |
+
"learning_rate": 1.2525972098545562e-07,
|
| 4397 |
+
"logits/chosen": 0.98828125,
|
| 4398 |
+
"logits/rejected": 1.1484375,
|
| 4399 |
+
"logps/chosen": -187.0,
|
| 4400 |
+
"logps/rejected": -330.0,
|
| 4401 |
+
"loss": 0.0,
|
| 4402 |
+
"rewards/accuracies": 1.0,
|
| 4403 |
+
"rewards/chosen": 3.640625,
|
| 4404 |
+
"rewards/margins": 22.5,
|
| 4405 |
+
"rewards/rejected": -18.875,
|
| 4406 |
+
"step": 2900
|
| 4407 |
+
},
|
| 4408 |
+
{
|
| 4409 |
+
"epoch": 2.331730769230769,
|
| 4410 |
+
"grad_norm": 1.0058948099975264e-05,
|
| 4411 |
+
"learning_rate": 1.2377560106856633e-07,
|
| 4412 |
+
"logits/chosen": 0.9140625,
|
| 4413 |
+
"logits/rejected": 1.296875,
|
| 4414 |
+
"logps/chosen": -146.0,
|
| 4415 |
+
"logps/rejected": -342.0,
|
| 4416 |
+
"loss": 0.0,
|
| 4417 |
+
"rewards/accuracies": 1.0,
|
| 4418 |
+
"rewards/chosen": 3.234375,
|
| 4419 |
+
"rewards/margins": 23.75,
|
| 4420 |
+
"rewards/rejected": -20.625,
|
| 4421 |
+
"step": 2910
|
| 4422 |
+
},
|
| 4423 |
+
{
|
| 4424 |
+
"epoch": 2.33974358974359,
|
| 4425 |
+
"grad_norm": 3.2768050524307317e-06,
|
| 4426 |
+
"learning_rate": 1.2229148115167704e-07,
|
| 4427 |
+
"logits/chosen": 0.87109375,
|
| 4428 |
+
"logits/rejected": 1.0,
|
| 4429 |
+
"logps/chosen": -145.0,
|
| 4430 |
+
"logps/rejected": -336.0,
|
| 4431 |
+
"loss": 0.0,
|
| 4432 |
+
"rewards/accuracies": 1.0,
|
| 4433 |
+
"rewards/chosen": 3.3125,
|
| 4434 |
+
"rewards/margins": 23.75,
|
| 4435 |
+
"rewards/rejected": -20.5,
|
| 4436 |
+
"step": 2920
|
| 4437 |
+
},
|
| 4438 |
+
{
|
| 4439 |
+
"epoch": 2.34775641025641,
|
| 4440 |
+
"grad_norm": 8.279411059572999e-06,
|
| 4441 |
+
"learning_rate": 1.2080736123478776e-07,
|
| 4442 |
+
"logits/chosen": 0.71484375,
|
| 4443 |
+
"logits/rejected": 1.296875,
|
| 4444 |
+
"logps/chosen": -183.0,
|
| 4445 |
+
"logps/rejected": -338.0,
|
| 4446 |
+
"loss": 0.0,
|
| 4447 |
+
"rewards/accuracies": 1.0,
|
| 4448 |
+
"rewards/chosen": 2.734375,
|
| 4449 |
+
"rewards/margins": 22.625,
|
| 4450 |
+
"rewards/rejected": -19.875,
|
| 4451 |
+
"step": 2930
|
| 4452 |
+
},
|
| 4453 |
+
{
|
| 4454 |
+
"epoch": 2.355769230769231,
|
| 4455 |
+
"grad_norm": 5.097273243887312e-06,
|
| 4456 |
+
"learning_rate": 1.1932324131789847e-07,
|
| 4457 |
+
"logits/chosen": 0.9375,
|
| 4458 |
+
"logits/rejected": 1.4453125,
|
| 4459 |
+
"logps/chosen": -161.0,
|
| 4460 |
+
"logps/rejected": -360.0,
|
| 4461 |
+
"loss": 0.0,
|
| 4462 |
+
"rewards/accuracies": 1.0,
|
| 4463 |
+
"rewards/chosen": 3.15625,
|
| 4464 |
+
"rewards/margins": 24.75,
|
| 4465 |
+
"rewards/rejected": -21.625,
|
| 4466 |
+
"step": 2940
|
| 4467 |
+
},
|
| 4468 |
+
{
|
| 4469 |
+
"epoch": 2.363782051282051,
|
| 4470 |
+
"grad_norm": 0.001364815117768273,
|
| 4471 |
+
"learning_rate": 1.178391214010092e-07,
|
| 4472 |
+
"logits/chosen": 0.82421875,
|
| 4473 |
+
"logits/rejected": 1.0546875,
|
| 4474 |
+
"logps/chosen": -196.0,
|
| 4475 |
+
"logps/rejected": -340.0,
|
| 4476 |
+
"loss": 0.0,
|
| 4477 |
+
"rewards/accuracies": 1.0,
|
| 4478 |
+
"rewards/chosen": 3.5,
|
| 4479 |
+
"rewards/margins": 22.5,
|
| 4480 |
+
"rewards/rejected": -18.875,
|
| 4481 |
+
"step": 2950
|
| 4482 |
+
},
|
| 4483 |
+
{
|
| 4484 |
+
"epoch": 2.371794871794872,
|
| 4485 |
+
"grad_norm": 0.00013766237352927737,
|
| 4486 |
+
"learning_rate": 1.1635500148411991e-07,
|
| 4487 |
+
"logits/chosen": 0.828125,
|
| 4488 |
+
"logits/rejected": 1.53125,
|
| 4489 |
+
"logps/chosen": -160.0,
|
| 4490 |
+
"logps/rejected": -324.0,
|
| 4491 |
+
"loss": 0.0,
|
| 4492 |
+
"rewards/accuracies": 1.0,
|
| 4493 |
+
"rewards/chosen": 3.46875,
|
| 4494 |
+
"rewards/margins": 22.875,
|
| 4495 |
+
"rewards/rejected": -19.5,
|
| 4496 |
+
"step": 2960
|
| 4497 |
+
},
|
| 4498 |
+
{
|
| 4499 |
+
"epoch": 2.3798076923076925,
|
| 4500 |
+
"grad_norm": 7.072038044334629e-06,
|
| 4501 |
+
"learning_rate": 1.1487088156723062e-07,
|
| 4502 |
+
"logits/chosen": 0.97265625,
|
| 4503 |
+
"logits/rejected": 1.4609375,
|
| 4504 |
+
"logps/chosen": -148.0,
|
| 4505 |
+
"logps/rejected": -318.0,
|
| 4506 |
+
"loss": 0.0,
|
| 4507 |
+
"rewards/accuracies": 1.0,
|
| 4508 |
+
"rewards/chosen": 3.1875,
|
| 4509 |
+
"rewards/margins": 22.75,
|
| 4510 |
+
"rewards/rejected": -19.625,
|
| 4511 |
+
"step": 2970
|
| 4512 |
+
},
|
| 4513 |
+
{
|
| 4514 |
+
"epoch": 2.3878205128205128,
|
| 4515 |
+
"grad_norm": 1.434535069439472e-06,
|
| 4516 |
+
"learning_rate": 1.1338676165034135e-07,
|
| 4517 |
+
"logits/chosen": 0.87890625,
|
| 4518 |
+
"logits/rejected": 0.93359375,
|
| 4519 |
+
"logps/chosen": -147.0,
|
| 4520 |
+
"logps/rejected": -342.0,
|
| 4521 |
+
"loss": 0.0002,
|
| 4522 |
+
"rewards/accuracies": 1.0,
|
| 4523 |
+
"rewards/chosen": 3.203125,
|
| 4524 |
+
"rewards/margins": 22.75,
|
| 4525 |
+
"rewards/rejected": -19.5,
|
| 4526 |
+
"step": 2980
|
| 4527 |
+
},
|
| 4528 |
+
{
|
| 4529 |
+
"epoch": 2.3958333333333335,
|
| 4530 |
+
"grad_norm": 0.0003150427679969826,
|
| 4531 |
+
"learning_rate": 1.1190264173345205e-07,
|
| 4532 |
+
"logits/chosen": 0.78125,
|
| 4533 |
+
"logits/rejected": 1.0703125,
|
| 4534 |
+
"logps/chosen": -185.0,
|
| 4535 |
+
"logps/rejected": -330.0,
|
| 4536 |
+
"loss": 0.0,
|
| 4537 |
+
"rewards/accuracies": 1.0,
|
| 4538 |
+
"rewards/chosen": 3.375,
|
| 4539 |
+
"rewards/margins": 22.875,
|
| 4540 |
+
"rewards/rejected": -19.5,
|
| 4541 |
+
"step": 2990
|
| 4542 |
+
},
|
| 4543 |
+
{
|
| 4544 |
+
"epoch": 2.4038461538461537,
|
| 4545 |
+
"grad_norm": 7.914823589207877e-05,
|
| 4546 |
+
"learning_rate": 1.1041852181656278e-07,
|
| 4547 |
+
"logits/chosen": 0.9296875,
|
| 4548 |
+
"logits/rejected": 1.421875,
|
| 4549 |
+
"logps/chosen": -151.0,
|
| 4550 |
+
"logps/rejected": -342.0,
|
| 4551 |
+
"loss": 0.0,
|
| 4552 |
+
"rewards/accuracies": 1.0,
|
| 4553 |
+
"rewards/chosen": 3.703125,
|
| 4554 |
+
"rewards/margins": 23.375,
|
| 4555 |
+
"rewards/rejected": -19.75,
|
| 4556 |
+
"step": 3000
|
| 4557 |
+
},
|
| 4558 |
+
{
|
| 4559 |
+
"epoch": 2.4118589743589745,
|
| 4560 |
+
"grad_norm": 0.00026905024279143275,
|
| 4561 |
+
"learning_rate": 1.0893440189967348e-07,
|
| 4562 |
+
"logits/chosen": 0.76171875,
|
| 4563 |
+
"logits/rejected": 1.34375,
|
| 4564 |
+
"logps/chosen": -161.0,
|
| 4565 |
+
"logps/rejected": -354.0,
|
| 4566 |
+
"loss": 0.0,
|
| 4567 |
+
"rewards/accuracies": 1.0,
|
| 4568 |
+
"rewards/chosen": 3.40625,
|
| 4569 |
+
"rewards/margins": 24.5,
|
| 4570 |
+
"rewards/rejected": -21.125,
|
| 4571 |
+
"step": 3010
|
| 4572 |
+
},
|
| 4573 |
+
{
|
| 4574 |
+
"epoch": 2.4198717948717947,
|
| 4575 |
+
"grad_norm": 2.5621741265848455e-06,
|
| 4576 |
+
"learning_rate": 1.074502819827842e-07,
|
| 4577 |
+
"logits/chosen": 0.83203125,
|
| 4578 |
+
"logits/rejected": 1.3515625,
|
| 4579 |
+
"logps/chosen": -155.0,
|
| 4580 |
+
"logps/rejected": -352.0,
|
| 4581 |
+
"loss": 0.0,
|
| 4582 |
+
"rewards/accuracies": 1.0,
|
| 4583 |
+
"rewards/chosen": 3.46875,
|
| 4584 |
+
"rewards/margins": 23.25,
|
| 4585 |
+
"rewards/rejected": -19.875,
|
| 4586 |
+
"step": 3020
|
| 4587 |
+
},
|
| 4588 |
+
{
|
| 4589 |
+
"epoch": 2.4278846153846154,
|
| 4590 |
+
"grad_norm": 0.0006476017033990356,
|
| 4591 |
+
"learning_rate": 1.0596616206589493e-07,
|
| 4592 |
+
"logits/chosen": 0.83203125,
|
| 4593 |
+
"logits/rejected": 1.3046875,
|
| 4594 |
+
"logps/chosen": -176.0,
|
| 4595 |
+
"logps/rejected": -326.0,
|
| 4596 |
+
"loss": 0.0,
|
| 4597 |
+
"rewards/accuracies": 1.0,
|
| 4598 |
+
"rewards/chosen": 3.03125,
|
| 4599 |
+
"rewards/margins": 22.0,
|
| 4600 |
+
"rewards/rejected": -19.0,
|
| 4601 |
+
"step": 3030
|
| 4602 |
+
},
|
| 4603 |
+
{
|
| 4604 |
+
"epoch": 2.435897435897436,
|
| 4605 |
+
"grad_norm": 1.1941928602219868e-05,
|
| 4606 |
+
"learning_rate": 1.0448204214900563e-07,
|
| 4607 |
+
"logits/chosen": 1.1171875,
|
| 4608 |
+
"logits/rejected": 1.125,
|
| 4609 |
+
"logps/chosen": -167.0,
|
| 4610 |
+
"logps/rejected": -336.0,
|
| 4611 |
+
"loss": 0.0,
|
| 4612 |
+
"rewards/accuracies": 1.0,
|
| 4613 |
+
"rewards/chosen": 3.21875,
|
| 4614 |
+
"rewards/margins": 22.75,
|
| 4615 |
+
"rewards/rejected": -19.5,
|
| 4616 |
+
"step": 3040
|
| 4617 |
+
},
|
| 4618 |
+
{
|
| 4619 |
+
"epoch": 2.4439102564102564,
|
| 4620 |
+
"grad_norm": 4.511019063462089e-06,
|
| 4621 |
+
"learning_rate": 1.0299792223211636e-07,
|
| 4622 |
+
"logits/chosen": 0.96875,
|
| 4623 |
+
"logits/rejected": 1.2421875,
|
| 4624 |
+
"logps/chosen": -173.0,
|
| 4625 |
+
"logps/rejected": -338.0,
|
| 4626 |
+
"loss": 0.0,
|
| 4627 |
+
"rewards/accuracies": 1.0,
|
| 4628 |
+
"rewards/chosen": 3.421875,
|
| 4629 |
+
"rewards/margins": 22.75,
|
| 4630 |
+
"rewards/rejected": -19.375,
|
| 4631 |
+
"step": 3050
|
| 4632 |
+
},
|
| 4633 |
+
{
|
| 4634 |
+
"epoch": 2.451923076923077,
|
| 4635 |
+
"grad_norm": 1.1909188473574983e-06,
|
| 4636 |
+
"learning_rate": 1.0151380231522706e-07,
|
| 4637 |
+
"logits/chosen": 0.71875,
|
| 4638 |
+
"logits/rejected": 1.1953125,
|
| 4639 |
+
"logps/chosen": -189.0,
|
| 4640 |
+
"logps/rejected": -338.0,
|
| 4641 |
+
"loss": 0.0,
|
| 4642 |
+
"rewards/accuracies": 1.0,
|
| 4643 |
+
"rewards/chosen": 3.703125,
|
| 4644 |
+
"rewards/margins": 23.875,
|
| 4645 |
+
"rewards/rejected": -20.125,
|
| 4646 |
+
"step": 3060
|
| 4647 |
+
},
|
| 4648 |
+
{
|
| 4649 |
+
"epoch": 2.4599358974358974,
|
| 4650 |
+
"grad_norm": 0.00016413461567918458,
|
| 4651 |
+
"learning_rate": 1.0002968239833778e-07,
|
| 4652 |
+
"logits/chosen": 0.8515625,
|
| 4653 |
+
"logits/rejected": 1.140625,
|
| 4654 |
+
"logps/chosen": -176.0,
|
| 4655 |
+
"logps/rejected": -360.0,
|
| 4656 |
+
"loss": 0.0,
|
| 4657 |
+
"rewards/accuracies": 1.0,
|
| 4658 |
+
"rewards/chosen": 3.046875,
|
| 4659 |
+
"rewards/margins": 24.0,
|
| 4660 |
+
"rewards/rejected": -21.0,
|
| 4661 |
+
"step": 3070
|
| 4662 |
+
},
|
| 4663 |
+
{
|
| 4664 |
+
"epoch": 2.467948717948718,
|
| 4665 |
+
"grad_norm": 0.0004207874699143046,
|
| 4666 |
+
"learning_rate": 9.854556248144849e-08,
|
| 4667 |
+
"logits/chosen": 1.0625,
|
| 4668 |
+
"logits/rejected": 1.0703125,
|
| 4669 |
+
"logps/chosen": -171.0,
|
| 4670 |
+
"logps/rejected": -336.0,
|
| 4671 |
+
"loss": 0.0,
|
| 4672 |
+
"rewards/accuracies": 1.0,
|
| 4673 |
+
"rewards/chosen": 3.265625,
|
| 4674 |
+
"rewards/margins": 23.0,
|
| 4675 |
+
"rewards/rejected": -19.75,
|
| 4676 |
+
"step": 3080
|
| 4677 |
+
},
|
| 4678 |
+
{
|
| 4679 |
+
"epoch": 2.4759615384615383,
|
| 4680 |
+
"grad_norm": 0.0013603810220500937,
|
| 4681 |
+
"learning_rate": 9.706144256455921e-08,
|
| 4682 |
+
"logits/chosen": 0.859375,
|
| 4683 |
+
"logits/rejected": 1.4453125,
|
| 4684 |
+
"logps/chosen": -201.0,
|
| 4685 |
+
"logps/rejected": -330.0,
|
| 4686 |
+
"loss": 0.0,
|
| 4687 |
+
"rewards/accuracies": 1.0,
|
| 4688 |
+
"rewards/chosen": 3.75,
|
| 4689 |
+
"rewards/margins": 23.0,
|
| 4690 |
+
"rewards/rejected": -19.25,
|
| 4691 |
+
"step": 3090
|
| 4692 |
+
},
|
| 4693 |
+
{
|
| 4694 |
+
"epoch": 2.483974358974359,
|
| 4695 |
+
"grad_norm": 1.8119361476861766e-07,
|
| 4696 |
+
"learning_rate": 9.557732264766994e-08,
|
| 4697 |
+
"logits/chosen": 1.125,
|
| 4698 |
+
"logits/rejected": 1.65625,
|
| 4699 |
+
"logps/chosen": -181.0,
|
| 4700 |
+
"logps/rejected": -320.0,
|
| 4701 |
+
"loss": 0.0,
|
| 4702 |
+
"rewards/accuracies": 1.0,
|
| 4703 |
+
"rewards/chosen": 2.859375,
|
| 4704 |
+
"rewards/margins": 21.875,
|
| 4705 |
+
"rewards/rejected": -19.0,
|
| 4706 |
+
"step": 3100
|
| 4707 |
+
},
|
| 4708 |
+
{
|
| 4709 |
+
"epoch": 2.4919871794871793,
|
| 4710 |
+
"grad_norm": 1.0849727745118161e-05,
|
| 4711 |
+
"learning_rate": 9.409320273078064e-08,
|
| 4712 |
+
"logits/chosen": 0.75,
|
| 4713 |
+
"logits/rejected": 1.25,
|
| 4714 |
+
"logps/chosen": -194.0,
|
| 4715 |
+
"logps/rejected": -346.0,
|
| 4716 |
+
"loss": 0.0001,
|
| 4717 |
+
"rewards/accuracies": 1.0,
|
| 4718 |
+
"rewards/chosen": 3.65625,
|
| 4719 |
+
"rewards/margins": 25.125,
|
| 4720 |
+
"rewards/rejected": -21.375,
|
| 4721 |
+
"step": 3110
|
| 4722 |
+
},
|
| 4723 |
+
{
|
| 4724 |
+
"epoch": 2.5,
|
| 4725 |
+
"grad_norm": 0.00190454125856642,
|
| 4726 |
+
"learning_rate": 9.260908281389137e-08,
|
| 4727 |
+
"logits/chosen": 0.79296875,
|
| 4728 |
+
"logits/rejected": 1.1328125,
|
| 4729 |
+
"logps/chosen": -174.0,
|
| 4730 |
+
"logps/rejected": -318.0,
|
| 4731 |
+
"loss": 0.0,
|
| 4732 |
+
"rewards/accuracies": 1.0,
|
| 4733 |
+
"rewards/chosen": 3.671875,
|
| 4734 |
+
"rewards/margins": 23.0,
|
| 4735 |
+
"rewards/rejected": -19.25,
|
| 4736 |
+
"step": 3120
|
| 4737 |
+
},
|
| 4738 |
+
{
|
| 4739 |
+
"epoch": 2.5080128205128203,
|
| 4740 |
+
"grad_norm": 4.817138155607492e-05,
|
| 4741 |
+
"learning_rate": 9.112496289700207e-08,
|
| 4742 |
+
"logits/chosen": 1.078125,
|
| 4743 |
+
"logits/rejected": 1.0546875,
|
| 4744 |
+
"logps/chosen": -151.0,
|
| 4745 |
+
"logps/rejected": -336.0,
|
| 4746 |
+
"loss": 0.0,
|
| 4747 |
+
"rewards/accuracies": 1.0,
|
| 4748 |
+
"rewards/chosen": 3.6875,
|
| 4749 |
+
"rewards/margins": 23.5,
|
| 4750 |
+
"rewards/rejected": -19.875,
|
| 4751 |
+
"step": 3130
|
| 4752 |
+
},
|
| 4753 |
+
{
|
| 4754 |
+
"epoch": 2.516025641025641,
|
| 4755 |
+
"grad_norm": 4.1634510560200736e-07,
|
| 4756 |
+
"learning_rate": 8.964084298011279e-08,
|
| 4757 |
+
"logits/chosen": 0.89453125,
|
| 4758 |
+
"logits/rejected": 1.21875,
|
| 4759 |
+
"logps/chosen": -165.0,
|
| 4760 |
+
"logps/rejected": -350.0,
|
| 4761 |
+
"loss": 0.0,
|
| 4762 |
+
"rewards/accuracies": 1.0,
|
| 4763 |
+
"rewards/chosen": 3.03125,
|
| 4764 |
+
"rewards/margins": 23.625,
|
| 4765 |
+
"rewards/rejected": -20.625,
|
| 4766 |
+
"step": 3140
|
| 4767 |
+
},
|
| 4768 |
+
{
|
| 4769 |
+
"epoch": 2.5240384615384617,
|
| 4770 |
+
"grad_norm": 0.0030510730836868943,
|
| 4771 |
+
"learning_rate": 8.815672306322349e-08,
|
| 4772 |
+
"logits/chosen": 0.83984375,
|
| 4773 |
+
"logits/rejected": 1.3359375,
|
| 4774 |
+
"logps/chosen": -150.0,
|
| 4775 |
+
"logps/rejected": -338.0,
|
| 4776 |
+
"loss": 0.0,
|
| 4777 |
+
"rewards/accuracies": 1.0,
|
| 4778 |
+
"rewards/chosen": 3.09375,
|
| 4779 |
+
"rewards/margins": 23.5,
|
| 4780 |
+
"rewards/rejected": -20.375,
|
| 4781 |
+
"step": 3150
|
| 4782 |
+
},
|
| 4783 |
+
{
|
| 4784 |
+
"epoch": 2.532051282051282,
|
| 4785 |
+
"grad_norm": 4.333397926948427e-05,
|
| 4786 |
+
"learning_rate": 8.667260314633422e-08,
|
| 4787 |
+
"logits/chosen": 0.8203125,
|
| 4788 |
+
"logits/rejected": 1.328125,
|
| 4789 |
+
"logps/chosen": -144.0,
|
| 4790 |
+
"logps/rejected": -350.0,
|
| 4791 |
+
"loss": 0.0,
|
| 4792 |
+
"rewards/accuracies": 1.0,
|
| 4793 |
+
"rewards/chosen": 3.65625,
|
| 4794 |
+
"rewards/margins": 24.625,
|
| 4795 |
+
"rewards/rejected": -21.0,
|
| 4796 |
+
"step": 3160
|
| 4797 |
+
},
|
| 4798 |
+
{
|
| 4799 |
+
"epoch": 2.5400641025641026,
|
| 4800 |
+
"grad_norm": 3.816937452899327e-05,
|
| 4801 |
+
"learning_rate": 8.518848322944495e-08,
|
| 4802 |
+
"logits/chosen": 0.83203125,
|
| 4803 |
+
"logits/rejected": 1.1953125,
|
| 4804 |
+
"logps/chosen": -184.0,
|
| 4805 |
+
"logps/rejected": -330.0,
|
| 4806 |
+
"loss": 0.0,
|
| 4807 |
+
"rewards/accuracies": 1.0,
|
| 4808 |
+
"rewards/chosen": 3.515625,
|
| 4809 |
+
"rewards/margins": 23.25,
|
| 4810 |
+
"rewards/rejected": -19.75,
|
| 4811 |
+
"step": 3170
|
| 4812 |
+
},
|
| 4813 |
+
{
|
| 4814 |
+
"epoch": 2.5480769230769234,
|
| 4815 |
+
"grad_norm": 4.438174995412161e-06,
|
| 4816 |
+
"learning_rate": 8.370436331255565e-08,
|
| 4817 |
+
"logits/chosen": 0.9921875,
|
| 4818 |
+
"logits/rejected": 1.234375,
|
| 4819 |
+
"logps/chosen": -137.0,
|
| 4820 |
+
"logps/rejected": -334.0,
|
| 4821 |
+
"loss": 0.0,
|
| 4822 |
+
"rewards/accuracies": 1.0,
|
| 4823 |
+
"rewards/chosen": 3.546875,
|
| 4824 |
+
"rewards/margins": 23.75,
|
| 4825 |
+
"rewards/rejected": -20.25,
|
| 4826 |
+
"step": 3180
|
| 4827 |
+
},
|
| 4828 |
+
{
|
| 4829 |
+
"epoch": 2.5560897435897436,
|
| 4830 |
+
"grad_norm": 1.2674210936494033e-05,
|
| 4831 |
+
"learning_rate": 8.222024339566637e-08,
|
| 4832 |
+
"logits/chosen": 0.9765625,
|
| 4833 |
+
"logits/rejected": 1.1953125,
|
| 4834 |
+
"logps/chosen": -189.0,
|
| 4835 |
+
"logps/rejected": -338.0,
|
| 4836 |
+
"loss": 0.0,
|
| 4837 |
+
"rewards/accuracies": 1.0,
|
| 4838 |
+
"rewards/chosen": 3.828125,
|
| 4839 |
+
"rewards/margins": 22.625,
|
| 4840 |
+
"rewards/rejected": -18.75,
|
| 4841 |
+
"step": 3190
|
| 4842 |
+
},
|
| 4843 |
+
{
|
| 4844 |
+
"epoch": 2.564102564102564,
|
| 4845 |
+
"grad_norm": 2.2470915894489486e-06,
|
| 4846 |
+
"learning_rate": 8.073612347877707e-08,
|
| 4847 |
+
"logits/chosen": 0.7578125,
|
| 4848 |
+
"logits/rejected": 0.94140625,
|
| 4849 |
+
"logps/chosen": -210.0,
|
| 4850 |
+
"logps/rejected": -336.0,
|
| 4851 |
+
"loss": 0.0,
|
| 4852 |
+
"rewards/accuracies": 1.0,
|
| 4853 |
+
"rewards/chosen": 2.8125,
|
| 4854 |
+
"rewards/margins": 22.875,
|
| 4855 |
+
"rewards/rejected": -20.0,
|
| 4856 |
+
"step": 3200
|
| 4857 |
+
},
|
| 4858 |
+
{
|
| 4859 |
+
"epoch": 2.5721153846153846,
|
| 4860 |
+
"grad_norm": 0.0013392786326408394,
|
| 4861 |
+
"learning_rate": 7.92520035618878e-08,
|
| 4862 |
+
"logits/chosen": 0.70703125,
|
| 4863 |
+
"logits/rejected": 1.0625,
|
| 4864 |
+
"logps/chosen": -166.0,
|
| 4865 |
+
"logps/rejected": -330.0,
|
| 4866 |
+
"loss": 0.0,
|
| 4867 |
+
"rewards/accuracies": 1.0,
|
| 4868 |
+
"rewards/chosen": 3.28125,
|
| 4869 |
+
"rewards/margins": 23.125,
|
| 4870 |
+
"rewards/rejected": -19.875,
|
| 4871 |
+
"step": 3210
|
| 4872 |
+
},
|
| 4873 |
+
{
|
| 4874 |
+
"epoch": 2.5801282051282053,
|
| 4875 |
+
"grad_norm": 1.1479055755141908e-06,
|
| 4876 |
+
"learning_rate": 7.776788364499851e-08,
|
| 4877 |
+
"logits/chosen": 0.392578125,
|
| 4878 |
+
"logits/rejected": 1.03125,
|
| 4879 |
+
"logps/chosen": -176.0,
|
| 4880 |
+
"logps/rejected": -342.0,
|
| 4881 |
+
"loss": 0.0,
|
| 4882 |
+
"rewards/accuracies": 1.0,
|
| 4883 |
+
"rewards/chosen": 3.3125,
|
| 4884 |
+
"rewards/margins": 23.75,
|
| 4885 |
+
"rewards/rejected": -20.5,
|
| 4886 |
+
"step": 3220
|
| 4887 |
+
},
|
| 4888 |
+
{
|
| 4889 |
+
"epoch": 2.5881410256410255,
|
| 4890 |
+
"grad_norm": 6.57721816711125e-08,
|
| 4891 |
+
"learning_rate": 7.628376372810923e-08,
|
| 4892 |
+
"logits/chosen": 0.859375,
|
| 4893 |
+
"logits/rejected": 0.92578125,
|
| 4894 |
+
"logps/chosen": -161.0,
|
| 4895 |
+
"logps/rejected": -348.0,
|
| 4896 |
+
"loss": 0.0,
|
| 4897 |
+
"rewards/accuracies": 1.0,
|
| 4898 |
+
"rewards/chosen": 3.3125,
|
| 4899 |
+
"rewards/margins": 25.5,
|
| 4900 |
+
"rewards/rejected": -22.25,
|
| 4901 |
+
"step": 3230
|
| 4902 |
+
},
|
| 4903 |
+
{
|
| 4904 |
+
"epoch": 2.5961538461538463,
|
| 4905 |
+
"grad_norm": 8.654157902362506e-07,
|
| 4906 |
+
"learning_rate": 7.479964381121995e-08,
|
| 4907 |
+
"logits/chosen": 0.84375,
|
| 4908 |
+
"logits/rejected": 1.34375,
|
| 4909 |
+
"logps/chosen": -131.0,
|
| 4910 |
+
"logps/rejected": -296.0,
|
| 4911 |
+
"loss": 0.0,
|
| 4912 |
+
"rewards/accuracies": 1.0,
|
| 4913 |
+
"rewards/chosen": 3.34375,
|
| 4914 |
+
"rewards/margins": 21.125,
|
| 4915 |
+
"rewards/rejected": -17.75,
|
| 4916 |
+
"step": 3240
|
| 4917 |
+
},
|
| 4918 |
+
{
|
| 4919 |
+
"epoch": 2.6041666666666665,
|
| 4920 |
+
"grad_norm": 2.107173176911769e-05,
|
| 4921 |
+
"learning_rate": 7.331552389433065e-08,
|
| 4922 |
+
"logits/chosen": 0.92578125,
|
| 4923 |
+
"logits/rejected": 1.2421875,
|
| 4924 |
+
"logps/chosen": -182.0,
|
| 4925 |
+
"logps/rejected": -332.0,
|
| 4926 |
+
"loss": 0.0,
|
| 4927 |
+
"rewards/accuracies": 1.0,
|
| 4928 |
+
"rewards/chosen": 3.75,
|
| 4929 |
+
"rewards/margins": 22.625,
|
| 4930 |
+
"rewards/rejected": -18.875,
|
| 4931 |
+
"step": 3250
|
| 4932 |
+
},
|
| 4933 |
+
{
|
| 4934 |
+
"epoch": 2.6121794871794872,
|
| 4935 |
+
"grad_norm": 3.0995378385019935e-05,
|
| 4936 |
+
"learning_rate": 7.183140397744138e-08,
|
| 4937 |
+
"logits/chosen": 0.78515625,
|
| 4938 |
+
"logits/rejected": 1.1484375,
|
| 4939 |
+
"logps/chosen": -185.0,
|
| 4940 |
+
"logps/rejected": -330.0,
|
| 4941 |
+
"loss": 0.0,
|
| 4942 |
+
"rewards/accuracies": 1.0,
|
| 4943 |
+
"rewards/chosen": 3.53125,
|
| 4944 |
+
"rewards/margins": 23.375,
|
| 4945 |
+
"rewards/rejected": -19.75,
|
| 4946 |
+
"step": 3260
|
| 4947 |
+
},
|
| 4948 |
+
{
|
| 4949 |
+
"epoch": 2.6201923076923075,
|
| 4950 |
+
"grad_norm": 7.163491430995317e-06,
|
| 4951 |
+
"learning_rate": 7.034728406055208e-08,
|
| 4952 |
+
"logits/chosen": 1.0390625,
|
| 4953 |
+
"logits/rejected": 1.03125,
|
| 4954 |
+
"logps/chosen": -163.0,
|
| 4955 |
+
"logps/rejected": -336.0,
|
| 4956 |
+
"loss": 0.0,
|
| 4957 |
+
"rewards/accuracies": 1.0,
|
| 4958 |
+
"rewards/chosen": 3.5,
|
| 4959 |
+
"rewards/margins": 23.625,
|
| 4960 |
+
"rewards/rejected": -20.125,
|
| 4961 |
+
"step": 3270
|
| 4962 |
+
},
|
| 4963 |
+
{
|
| 4964 |
+
"epoch": 2.628205128205128,
|
| 4965 |
+
"grad_norm": 3.2825173935148675e-06,
|
| 4966 |
+
"learning_rate": 6.886316414366281e-08,
|
| 4967 |
+
"logits/chosen": 0.9921875,
|
| 4968 |
+
"logits/rejected": 1.1875,
|
| 4969 |
+
"logps/chosen": -167.0,
|
| 4970 |
+
"logps/rejected": -346.0,
|
| 4971 |
+
"loss": 0.0,
|
| 4972 |
+
"rewards/accuracies": 1.0,
|
| 4973 |
+
"rewards/chosen": 3.578125,
|
| 4974 |
+
"rewards/margins": 23.875,
|
| 4975 |
+
"rewards/rejected": -20.375,
|
| 4976 |
+
"step": 3280
|
| 4977 |
+
},
|
| 4978 |
+
{
|
| 4979 |
+
"epoch": 2.636217948717949,
|
| 4980 |
+
"grad_norm": 3.7018247685930666e-06,
|
| 4981 |
+
"learning_rate": 6.737904422677352e-08,
|
| 4982 |
+
"logits/chosen": 0.64453125,
|
| 4983 |
+
"logits/rejected": 1.125,
|
| 4984 |
+
"logps/chosen": -154.0,
|
| 4985 |
+
"logps/rejected": -340.0,
|
| 4986 |
+
"loss": 0.0,
|
| 4987 |
+
"rewards/accuracies": 1.0,
|
| 4988 |
+
"rewards/chosen": 3.28125,
|
| 4989 |
+
"rewards/margins": 25.0,
|
| 4990 |
+
"rewards/rejected": -21.625,
|
| 4991 |
+
"step": 3290
|
| 4992 |
+
},
|
| 4993 |
+
{
|
| 4994 |
+
"epoch": 2.644230769230769,
|
| 4995 |
+
"grad_norm": 1.5235500032868652e-06,
|
| 4996 |
+
"learning_rate": 6.589492430988424e-08,
|
| 4997 |
+
"logits/chosen": 0.86328125,
|
| 4998 |
+
"logits/rejected": 1.171875,
|
| 4999 |
+
"logps/chosen": -158.0,
|
| 5000 |
+
"logps/rejected": -326.0,
|
| 5001 |
+
"loss": 0.0,
|
| 5002 |
+
"rewards/accuracies": 1.0,
|
| 5003 |
+
"rewards/chosen": 4.0625,
|
| 5004 |
+
"rewards/margins": 23.0,
|
| 5005 |
+
"rewards/rejected": -19.0,
|
| 5006 |
+
"step": 3300
|
| 5007 |
+
},
|
| 5008 |
+
{
|
| 5009 |
+
"epoch": 2.65224358974359,
|
| 5010 |
+
"grad_norm": 2.4028100208495416e-05,
|
| 5011 |
+
"learning_rate": 6.441080439299495e-08,
|
| 5012 |
+
"logits/chosen": 0.9453125,
|
| 5013 |
+
"logits/rejected": 1.203125,
|
| 5014 |
+
"logps/chosen": -174.0,
|
| 5015 |
+
"logps/rejected": -330.0,
|
| 5016 |
+
"loss": 0.0,
|
| 5017 |
+
"rewards/accuracies": 1.0,
|
| 5018 |
+
"rewards/chosen": 3.484375,
|
| 5019 |
+
"rewards/margins": 23.375,
|
| 5020 |
+
"rewards/rejected": -19.875,
|
| 5021 |
+
"step": 3310
|
| 5022 |
+
},
|
| 5023 |
+
{
|
| 5024 |
+
"epoch": 2.66025641025641,
|
| 5025 |
+
"grad_norm": 0.0006641311306163846,
|
| 5026 |
+
"learning_rate": 6.292668447610566e-08,
|
| 5027 |
+
"logits/chosen": 1.03125,
|
| 5028 |
+
"logits/rejected": 1.375,
|
| 5029 |
+
"logps/chosen": -154.0,
|
| 5030 |
+
"logps/rejected": -348.0,
|
| 5031 |
+
"loss": 0.0002,
|
| 5032 |
+
"rewards/accuracies": 1.0,
|
| 5033 |
+
"rewards/chosen": 3.671875,
|
| 5034 |
+
"rewards/margins": 24.25,
|
| 5035 |
+
"rewards/rejected": -20.5,
|
| 5036 |
+
"step": 3320
|
| 5037 |
+
},
|
| 5038 |
+
{
|
| 5039 |
+
"epoch": 2.668269230769231,
|
| 5040 |
+
"grad_norm": 1.3261677467269389e-05,
|
| 5041 |
+
"learning_rate": 6.144256455921639e-08,
|
| 5042 |
+
"logits/chosen": 1.171875,
|
| 5043 |
+
"logits/rejected": 1.671875,
|
| 5044 |
+
"logps/chosen": -136.0,
|
| 5045 |
+
"logps/rejected": -342.0,
|
| 5046 |
+
"loss": 0.0,
|
| 5047 |
+
"rewards/accuracies": 1.0,
|
| 5048 |
+
"rewards/chosen": 3.6875,
|
| 5049 |
+
"rewards/margins": 24.0,
|
| 5050 |
+
"rewards/rejected": -20.375,
|
| 5051 |
+
"step": 3330
|
| 5052 |
+
},
|
| 5053 |
+
{
|
| 5054 |
+
"epoch": 2.676282051282051,
|
| 5055 |
+
"grad_norm": 2.3998995719683365e-07,
|
| 5056 |
+
"learning_rate": 5.99584446423271e-08,
|
| 5057 |
+
"logits/chosen": 0.890625,
|
| 5058 |
+
"logits/rejected": 1.2890625,
|
| 5059 |
+
"logps/chosen": -212.0,
|
| 5060 |
+
"logps/rejected": -320.0,
|
| 5061 |
+
"loss": 0.0,
|
| 5062 |
+
"rewards/accuracies": 1.0,
|
| 5063 |
+
"rewards/chosen": 3.40625,
|
| 5064 |
+
"rewards/margins": 22.875,
|
| 5065 |
+
"rewards/rejected": -19.5,
|
| 5066 |
+
"step": 3340
|
| 5067 |
+
},
|
| 5068 |
+
{
|
| 5069 |
+
"epoch": 2.684294871794872,
|
| 5070 |
+
"grad_norm": 1.1770292696948012e-07,
|
| 5071 |
+
"learning_rate": 5.847432472543781e-08,
|
| 5072 |
+
"logits/chosen": 0.83984375,
|
| 5073 |
+
"logits/rejected": 1.6484375,
|
| 5074 |
+
"logps/chosen": -168.0,
|
| 5075 |
+
"logps/rejected": -342.0,
|
| 5076 |
+
"loss": 0.0,
|
| 5077 |
+
"rewards/accuracies": 1.0,
|
| 5078 |
+
"rewards/chosen": 3.484375,
|
| 5079 |
+
"rewards/margins": 24.5,
|
| 5080 |
+
"rewards/rejected": -21.0,
|
| 5081 |
+
"step": 3350
|
| 5082 |
+
},
|
| 5083 |
+
{
|
| 5084 |
+
"epoch": 2.6923076923076925,
|
| 5085 |
+
"grad_norm": 2.900580842341117e-07,
|
| 5086 |
+
"learning_rate": 5.699020480854853e-08,
|
| 5087 |
+
"logits/chosen": 0.91796875,
|
| 5088 |
+
"logits/rejected": 1.09375,
|
| 5089 |
+
"logps/chosen": -142.0,
|
| 5090 |
+
"logps/rejected": -352.0,
|
| 5091 |
+
"loss": 0.0,
|
| 5092 |
+
"rewards/accuracies": 1.0,
|
| 5093 |
+
"rewards/chosen": 3.53125,
|
| 5094 |
+
"rewards/margins": 25.625,
|
| 5095 |
+
"rewards/rejected": -22.125,
|
| 5096 |
+
"step": 3360
|
| 5097 |
+
},
|
| 5098 |
+
{
|
| 5099 |
+
"epoch": 2.7003205128205128,
|
| 5100 |
+
"grad_norm": 1.4745279648366565e-05,
|
| 5101 |
+
"learning_rate": 5.550608489165924e-08,
|
| 5102 |
+
"logits/chosen": 0.84375,
|
| 5103 |
+
"logits/rejected": 1.09375,
|
| 5104 |
+
"logps/chosen": -150.0,
|
| 5105 |
+
"logps/rejected": -324.0,
|
| 5106 |
+
"loss": 0.0,
|
| 5107 |
+
"rewards/accuracies": 1.0,
|
| 5108 |
+
"rewards/chosen": 3.578125,
|
| 5109 |
+
"rewards/margins": 23.25,
|
| 5110 |
+
"rewards/rejected": -19.75,
|
| 5111 |
+
"step": 3370
|
| 5112 |
+
},
|
| 5113 |
+
{
|
| 5114 |
+
"epoch": 2.7083333333333335,
|
| 5115 |
+
"grad_norm": 1.4514272796826e-06,
|
| 5116 |
+
"learning_rate": 5.4021964974769963e-08,
|
| 5117 |
+
"logits/chosen": 0.92578125,
|
| 5118 |
+
"logits/rejected": 1.359375,
|
| 5119 |
+
"logps/chosen": -163.0,
|
| 5120 |
+
"logps/rejected": -332.0,
|
| 5121 |
+
"loss": 0.0,
|
| 5122 |
+
"rewards/accuracies": 1.0,
|
| 5123 |
+
"rewards/chosen": 3.375,
|
| 5124 |
+
"rewards/margins": 23.0,
|
| 5125 |
+
"rewards/rejected": -19.625,
|
| 5126 |
+
"step": 3380
|
| 5127 |
+
},
|
| 5128 |
+
{
|
| 5129 |
+
"epoch": 2.7163461538461537,
|
| 5130 |
+
"grad_norm": 0.0005888690086543229,
|
| 5131 |
+
"learning_rate": 5.253784505788068e-08,
|
| 5132 |
+
"logits/chosen": 1.03125,
|
| 5133 |
+
"logits/rejected": 1.0234375,
|
| 5134 |
+
"logps/chosen": -155.0,
|
| 5135 |
+
"logps/rejected": -328.0,
|
| 5136 |
+
"loss": 0.0,
|
| 5137 |
+
"rewards/accuracies": 1.0,
|
| 5138 |
+
"rewards/chosen": 3.125,
|
| 5139 |
+
"rewards/margins": 22.625,
|
| 5140 |
+
"rewards/rejected": -19.5,
|
| 5141 |
+
"step": 3390
|
| 5142 |
+
},
|
| 5143 |
+
{
|
| 5144 |
+
"epoch": 2.7243589743589745,
|
| 5145 |
+
"grad_norm": 2.0042622885174553e-05,
|
| 5146 |
+
"learning_rate": 5.105372514099139e-08,
|
| 5147 |
+
"logits/chosen": 0.87890625,
|
| 5148 |
+
"logits/rejected": 1.390625,
|
| 5149 |
+
"logps/chosen": -144.0,
|
| 5150 |
+
"logps/rejected": -336.0,
|
| 5151 |
+
"loss": 0.0,
|
| 5152 |
+
"rewards/accuracies": 1.0,
|
| 5153 |
+
"rewards/chosen": 3.6875,
|
| 5154 |
+
"rewards/margins": 23.5,
|
| 5155 |
+
"rewards/rejected": -19.875,
|
| 5156 |
+
"step": 3400
|
| 5157 |
+
},
|
| 5158 |
+
{
|
| 5159 |
+
"epoch": 2.7323717948717947,
|
| 5160 |
+
"grad_norm": 5.635967172397906e-05,
|
| 5161 |
+
"learning_rate": 4.9569605224102104e-08,
|
| 5162 |
+
"logits/chosen": 0.99609375,
|
| 5163 |
+
"logits/rejected": 1.4375,
|
| 5164 |
+
"logps/chosen": -171.0,
|
| 5165 |
+
"logps/rejected": -336.0,
|
| 5166 |
+
"loss": 0.0,
|
| 5167 |
+
"rewards/accuracies": 1.0,
|
| 5168 |
+
"rewards/chosen": 3.546875,
|
| 5169 |
+
"rewards/margins": 24.125,
|
| 5170 |
+
"rewards/rejected": -20.625,
|
| 5171 |
+
"step": 3410
|
| 5172 |
+
},
|
| 5173 |
+
{
|
| 5174 |
+
"epoch": 2.7403846153846154,
|
| 5175 |
+
"grad_norm": 6.551150056642263e-06,
|
| 5176 |
+
"learning_rate": 4.808548530721282e-08,
|
| 5177 |
+
"logits/chosen": 1.0078125,
|
| 5178 |
+
"logits/rejected": 1.359375,
|
| 5179 |
+
"logps/chosen": -172.0,
|
| 5180 |
+
"logps/rejected": -350.0,
|
| 5181 |
+
"loss": 0.0,
|
| 5182 |
+
"rewards/accuracies": 1.0,
|
| 5183 |
+
"rewards/chosen": 3.609375,
|
| 5184 |
+
"rewards/margins": 24.5,
|
| 5185 |
+
"rewards/rejected": -20.875,
|
| 5186 |
+
"step": 3420
|
| 5187 |
+
},
|
| 5188 |
+
{
|
| 5189 |
+
"epoch": 2.748397435897436,
|
| 5190 |
+
"grad_norm": 2.170737569638303e-05,
|
| 5191 |
+
"learning_rate": 4.660136539032353e-08,
|
| 5192 |
+
"logits/chosen": 1.09375,
|
| 5193 |
+
"logits/rejected": 1.3203125,
|
| 5194 |
+
"logps/chosen": -170.0,
|
| 5195 |
+
"logps/rejected": -348.0,
|
| 5196 |
+
"loss": 0.0,
|
| 5197 |
+
"rewards/accuracies": 1.0,
|
| 5198 |
+
"rewards/chosen": 3.40625,
|
| 5199 |
+
"rewards/margins": 24.75,
|
| 5200 |
+
"rewards/rejected": -21.375,
|
| 5201 |
+
"step": 3430
|
| 5202 |
+
},
|
| 5203 |
+
{
|
| 5204 |
+
"epoch": 2.7564102564102564,
|
| 5205 |
+
"grad_norm": 3.662218737713611e-05,
|
| 5206 |
+
"learning_rate": 4.511724547343425e-08,
|
| 5207 |
+
"logits/chosen": 0.734375,
|
| 5208 |
+
"logits/rejected": 1.3671875,
|
| 5209 |
+
"logps/chosen": -139.0,
|
| 5210 |
+
"logps/rejected": -334.0,
|
| 5211 |
+
"loss": 0.0,
|
| 5212 |
+
"rewards/accuracies": 1.0,
|
| 5213 |
+
"rewards/chosen": 3.359375,
|
| 5214 |
+
"rewards/margins": 23.5,
|
| 5215 |
+
"rewards/rejected": -20.125,
|
| 5216 |
+
"step": 3440
|
| 5217 |
+
},
|
| 5218 |
+
{
|
| 5219 |
+
"epoch": 2.7644230769230766,
|
| 5220 |
+
"grad_norm": 2.0092823057660767e-07,
|
| 5221 |
+
"learning_rate": 4.363312555654497e-08,
|
| 5222 |
+
"logits/chosen": 1.0078125,
|
| 5223 |
+
"logits/rejected": 1.1640625,
|
| 5224 |
+
"logps/chosen": -178.0,
|
| 5225 |
+
"logps/rejected": -334.0,
|
| 5226 |
+
"loss": 0.0,
|
| 5227 |
+
"rewards/accuracies": 1.0,
|
| 5228 |
+
"rewards/chosen": 3.53125,
|
| 5229 |
+
"rewards/margins": 22.5,
|
| 5230 |
+
"rewards/rejected": -19.0,
|
| 5231 |
+
"step": 3450
|
| 5232 |
+
},
|
| 5233 |
+
{
|
| 5234 |
+
"epoch": 2.7724358974358974,
|
| 5235 |
+
"grad_norm": 8.047688007594987e-05,
|
| 5236 |
+
"learning_rate": 4.2149005639655685e-08,
|
| 5237 |
+
"logits/chosen": 0.8046875,
|
| 5238 |
+
"logits/rejected": 1.2578125,
|
| 5239 |
+
"logps/chosen": -163.0,
|
| 5240 |
+
"logps/rejected": -354.0,
|
| 5241 |
+
"loss": 0.0,
|
| 5242 |
+
"rewards/accuracies": 1.0,
|
| 5243 |
+
"rewards/chosen": 3.46875,
|
| 5244 |
+
"rewards/margins": 24.75,
|
| 5245 |
+
"rewards/rejected": -21.25,
|
| 5246 |
+
"step": 3460
|
| 5247 |
+
},
|
| 5248 |
+
{
|
| 5249 |
+
"epoch": 2.780448717948718,
|
| 5250 |
+
"grad_norm": 0.0006711480616614469,
|
| 5251 |
+
"learning_rate": 4.06648857227664e-08,
|
| 5252 |
+
"logits/chosen": 0.76953125,
|
| 5253 |
+
"logits/rejected": 1.40625,
|
| 5254 |
+
"logps/chosen": -167.0,
|
| 5255 |
+
"logps/rejected": -330.0,
|
| 5256 |
+
"loss": 0.0,
|
| 5257 |
+
"rewards/accuracies": 1.0,
|
| 5258 |
+
"rewards/chosen": 3.421875,
|
| 5259 |
+
"rewards/margins": 23.125,
|
| 5260 |
+
"rewards/rejected": -19.75,
|
| 5261 |
+
"step": 3470
|
| 5262 |
+
},
|
| 5263 |
+
{
|
| 5264 |
+
"epoch": 2.7884615384615383,
|
| 5265 |
+
"grad_norm": 0.0003280497102533483,
|
| 5266 |
+
"learning_rate": 3.918076580587711e-08,
|
| 5267 |
+
"logits/chosen": 0.6875,
|
| 5268 |
+
"logits/rejected": 1.1796875,
|
| 5269 |
+
"logps/chosen": -174.0,
|
| 5270 |
+
"logps/rejected": -350.0,
|
| 5271 |
+
"loss": 0.0,
|
| 5272 |
+
"rewards/accuracies": 1.0,
|
| 5273 |
+
"rewards/chosen": 3.265625,
|
| 5274 |
+
"rewards/margins": 23.125,
|
| 5275 |
+
"rewards/rejected": -19.875,
|
| 5276 |
+
"step": 3480
|
| 5277 |
+
},
|
| 5278 |
+
{
|
| 5279 |
+
"epoch": 2.796474358974359,
|
| 5280 |
+
"grad_norm": 6.798795586197906e-05,
|
| 5281 |
+
"learning_rate": 3.7696645888987825e-08,
|
| 5282 |
+
"logits/chosen": 0.87109375,
|
| 5283 |
+
"logits/rejected": 0.99609375,
|
| 5284 |
+
"logps/chosen": -147.0,
|
| 5285 |
+
"logps/rejected": -356.0,
|
| 5286 |
+
"loss": 0.0,
|
| 5287 |
+
"rewards/accuracies": 1.0,
|
| 5288 |
+
"rewards/chosen": 3.265625,
|
| 5289 |
+
"rewards/margins": 24.625,
|
| 5290 |
+
"rewards/rejected": -21.375,
|
| 5291 |
+
"step": 3490
|
| 5292 |
+
},
|
| 5293 |
+
{
|
| 5294 |
+
"epoch": 2.8044871794871797,
|
| 5295 |
+
"grad_norm": 4.364786537805849e-06,
|
| 5296 |
+
"learning_rate": 3.621252597209854e-08,
|
| 5297 |
+
"logits/chosen": 0.91796875,
|
| 5298 |
+
"logits/rejected": 1.3984375,
|
| 5299 |
+
"logps/chosen": -159.0,
|
| 5300 |
+
"logps/rejected": -338.0,
|
| 5301 |
+
"loss": 0.0,
|
| 5302 |
+
"rewards/accuracies": 1.0,
|
| 5303 |
+
"rewards/chosen": 3.375,
|
| 5304 |
+
"rewards/margins": 23.625,
|
| 5305 |
+
"rewards/rejected": -20.25,
|
| 5306 |
+
"step": 3500
|
| 5307 |
+
},
|
| 5308 |
+
{
|
| 5309 |
+
"epoch": 2.8125,
|
| 5310 |
+
"grad_norm": 1.9387537093442454e-08,
|
| 5311 |
+
"learning_rate": 3.4728406055209265e-08,
|
| 5312 |
+
"logits/chosen": 0.75390625,
|
| 5313 |
+
"logits/rejected": 1.1171875,
|
| 5314 |
+
"logps/chosen": -185.0,
|
| 5315 |
+
"logps/rejected": -348.0,
|
| 5316 |
+
"loss": 0.0,
|
| 5317 |
+
"rewards/accuracies": 1.0,
|
| 5318 |
+
"rewards/chosen": 3.578125,
|
| 5319 |
+
"rewards/margins": 25.125,
|
| 5320 |
+
"rewards/rejected": -21.5,
|
| 5321 |
+
"step": 3510
|
| 5322 |
+
},
|
| 5323 |
+
{
|
| 5324 |
+
"epoch": 2.8205128205128203,
|
| 5325 |
+
"grad_norm": 1.3880341861407807e-05,
|
| 5326 |
+
"learning_rate": 3.324428613831998e-08,
|
| 5327 |
+
"logits/chosen": 1.015625,
|
| 5328 |
+
"logits/rejected": 1.296875,
|
| 5329 |
+
"logps/chosen": -143.0,
|
| 5330 |
+
"logps/rejected": -358.0,
|
| 5331 |
+
"loss": 0.0,
|
| 5332 |
+
"rewards/accuracies": 1.0,
|
| 5333 |
+
"rewards/chosen": 3.53125,
|
| 5334 |
+
"rewards/margins": 24.75,
|
| 5335 |
+
"rewards/rejected": -21.25,
|
| 5336 |
+
"step": 3520
|
| 5337 |
+
},
|
| 5338 |
+
{
|
| 5339 |
+
"epoch": 2.828525641025641,
|
| 5340 |
+
"grad_norm": 1.8304365557650966e-05,
|
| 5341 |
+
"learning_rate": 3.176016622143069e-08,
|
| 5342 |
+
"logits/chosen": 1.046875,
|
| 5343 |
+
"logits/rejected": 1.203125,
|
| 5344 |
+
"logps/chosen": -167.0,
|
| 5345 |
+
"logps/rejected": -340.0,
|
| 5346 |
+
"loss": 0.0,
|
| 5347 |
+
"rewards/accuracies": 1.0,
|
| 5348 |
+
"rewards/chosen": 3.78125,
|
| 5349 |
+
"rewards/margins": 24.25,
|
| 5350 |
+
"rewards/rejected": -20.5,
|
| 5351 |
+
"step": 3530
|
| 5352 |
+
},
|
| 5353 |
+
{
|
| 5354 |
+
"epoch": 2.8365384615384617,
|
| 5355 |
+
"grad_norm": 5.07839088524003e-07,
|
| 5356 |
+
"learning_rate": 3.0276046304541406e-08,
|
| 5357 |
+
"logits/chosen": 0.94140625,
|
| 5358 |
+
"logits/rejected": 1.203125,
|
| 5359 |
+
"logps/chosen": -170.0,
|
| 5360 |
+
"logps/rejected": -344.0,
|
| 5361 |
+
"loss": 0.0,
|
| 5362 |
+
"rewards/accuracies": 1.0,
|
| 5363 |
+
"rewards/chosen": 3.140625,
|
| 5364 |
+
"rewards/margins": 23.75,
|
| 5365 |
+
"rewards/rejected": -20.625,
|
| 5366 |
+
"step": 3540
|
| 5367 |
+
},
|
| 5368 |
+
{
|
| 5369 |
+
"epoch": 2.844551282051282,
|
| 5370 |
+
"grad_norm": 6.497502370566641e-06,
|
| 5371 |
+
"learning_rate": 2.879192638765212e-08,
|
| 5372 |
+
"logits/chosen": 0.7578125,
|
| 5373 |
+
"logits/rejected": 1.09375,
|
| 5374 |
+
"logps/chosen": -180.0,
|
| 5375 |
+
"logps/rejected": -320.0,
|
| 5376 |
+
"loss": 0.0,
|
| 5377 |
+
"rewards/accuracies": 1.0,
|
| 5378 |
+
"rewards/chosen": 3.796875,
|
| 5379 |
+
"rewards/margins": 22.125,
|
| 5380 |
+
"rewards/rejected": -18.375,
|
| 5381 |
+
"step": 3550
|
| 5382 |
+
},
|
| 5383 |
+
{
|
| 5384 |
+
"epoch": 2.8525641025641026,
|
| 5385 |
+
"grad_norm": 2.0919183207981193e-05,
|
| 5386 |
+
"learning_rate": 2.7307806470762836e-08,
|
| 5387 |
+
"logits/chosen": 0.93359375,
|
| 5388 |
+
"logits/rejected": 1.4609375,
|
| 5389 |
+
"logps/chosen": -171.0,
|
| 5390 |
+
"logps/rejected": -334.0,
|
| 5391 |
+
"loss": 0.0,
|
| 5392 |
+
"rewards/accuracies": 1.0,
|
| 5393 |
+
"rewards/chosen": 3.65625,
|
| 5394 |
+
"rewards/margins": 23.25,
|
| 5395 |
+
"rewards/rejected": -19.5,
|
| 5396 |
+
"step": 3560
|
| 5397 |
+
},
|
| 5398 |
+
{
|
| 5399 |
+
"epoch": 2.8605769230769234,
|
| 5400 |
+
"grad_norm": 5.495725650489294e-08,
|
| 5401 |
+
"learning_rate": 2.5823686553873553e-08,
|
| 5402 |
+
"logits/chosen": 0.8359375,
|
| 5403 |
+
"logits/rejected": 1.140625,
|
| 5404 |
+
"logps/chosen": -166.0,
|
| 5405 |
+
"logps/rejected": -366.0,
|
| 5406 |
+
"loss": 0.0,
|
| 5407 |
+
"rewards/accuracies": 1.0,
|
| 5408 |
+
"rewards/chosen": 2.984375,
|
| 5409 |
+
"rewards/margins": 25.125,
|
| 5410 |
+
"rewards/rejected": -22.125,
|
| 5411 |
+
"step": 3570
|
| 5412 |
+
},
|
| 5413 |
+
{
|
| 5414 |
+
"epoch": 2.8685897435897436,
|
| 5415 |
+
"grad_norm": 0.00027587029209878266,
|
| 5416 |
+
"learning_rate": 2.4339566636984267e-08,
|
| 5417 |
+
"logits/chosen": 0.92578125,
|
| 5418 |
+
"logits/rejected": 1.46875,
|
| 5419 |
+
"logps/chosen": -178.0,
|
| 5420 |
+
"logps/rejected": -314.0,
|
| 5421 |
+
"loss": 0.0,
|
| 5422 |
+
"rewards/accuracies": 1.0,
|
| 5423 |
+
"rewards/chosen": 3.78125,
|
| 5424 |
+
"rewards/margins": 21.75,
|
| 5425 |
+
"rewards/rejected": -18.0,
|
| 5426 |
+
"step": 3580
|
| 5427 |
+
},
|
| 5428 |
+
{
|
| 5429 |
+
"epoch": 2.876602564102564,
|
| 5430 |
+
"grad_norm": 3.145906163646421e-06,
|
| 5431 |
+
"learning_rate": 2.2855446720094983e-08,
|
| 5432 |
+
"logits/chosen": 0.93359375,
|
| 5433 |
+
"logits/rejected": 1.2109375,
|
| 5434 |
+
"logps/chosen": -168.0,
|
| 5435 |
+
"logps/rejected": -356.0,
|
| 5436 |
+
"loss": 0.0,
|
| 5437 |
+
"rewards/accuracies": 1.0,
|
| 5438 |
+
"rewards/chosen": 3.390625,
|
| 5439 |
+
"rewards/margins": 24.875,
|
| 5440 |
+
"rewards/rejected": -21.5,
|
| 5441 |
+
"step": 3590
|
| 5442 |
+
},
|
| 5443 |
+
{
|
| 5444 |
+
"epoch": 2.8846153846153846,
|
| 5445 |
+
"grad_norm": 0.002758379167699333,
|
| 5446 |
+
"learning_rate": 2.1371326803205697e-08,
|
| 5447 |
+
"logits/chosen": 0.92578125,
|
| 5448 |
+
"logits/rejected": 1.46875,
|
| 5449 |
+
"logps/chosen": -172.0,
|
| 5450 |
+
"logps/rejected": -328.0,
|
| 5451 |
+
"loss": 0.0,
|
| 5452 |
+
"rewards/accuracies": 1.0,
|
| 5453 |
+
"rewards/chosen": 3.875,
|
| 5454 |
+
"rewards/margins": 23.25,
|
| 5455 |
+
"rewards/rejected": -19.375,
|
| 5456 |
+
"step": 3600
|
| 5457 |
+
},
|
| 5458 |
+
{
|
| 5459 |
+
"epoch": 2.8926282051282053,
|
| 5460 |
+
"grad_norm": 3.8764981035982807e-07,
|
| 5461 |
+
"learning_rate": 1.9887206886316414e-08,
|
| 5462 |
+
"logits/chosen": 0.8046875,
|
| 5463 |
+
"logits/rejected": 1.4375,
|
| 5464 |
+
"logps/chosen": -166.0,
|
| 5465 |
+
"logps/rejected": -364.0,
|
| 5466 |
+
"loss": 0.0,
|
| 5467 |
+
"rewards/accuracies": 1.0,
|
| 5468 |
+
"rewards/chosen": 2.875,
|
| 5469 |
+
"rewards/margins": 25.25,
|
| 5470 |
+
"rewards/rejected": -22.5,
|
| 5471 |
+
"step": 3610
|
| 5472 |
+
},
|
| 5473 |
+
{
|
| 5474 |
+
"epoch": 2.9006410256410255,
|
| 5475 |
+
"grad_norm": 2.459298995862145e-05,
|
| 5476 |
+
"learning_rate": 1.840308696942713e-08,
|
| 5477 |
+
"logits/chosen": 0.9296875,
|
| 5478 |
+
"logits/rejected": 1.2109375,
|
| 5479 |
+
"logps/chosen": -170.0,
|
| 5480 |
+
"logps/rejected": -346.0,
|
| 5481 |
+
"loss": 0.0,
|
| 5482 |
+
"rewards/accuracies": 1.0,
|
| 5483 |
+
"rewards/chosen": 3.390625,
|
| 5484 |
+
"rewards/margins": 24.0,
|
| 5485 |
+
"rewards/rejected": -20.5,
|
| 5486 |
+
"step": 3620
|
| 5487 |
+
},
|
| 5488 |
+
{
|
| 5489 |
+
"epoch": 2.9086538461538463,
|
| 5490 |
+
"grad_norm": 0.0033663621977632068,
|
| 5491 |
+
"learning_rate": 1.6918967052537844e-08,
|
| 5492 |
+
"logits/chosen": 0.90234375,
|
| 5493 |
+
"logits/rejected": 1.1015625,
|
| 5494 |
+
"logps/chosen": -158.0,
|
| 5495 |
+
"logps/rejected": -342.0,
|
| 5496 |
+
"loss": 0.0,
|
| 5497 |
+
"rewards/accuracies": 1.0,
|
| 5498 |
+
"rewards/chosen": 3.640625,
|
| 5499 |
+
"rewards/margins": 24.625,
|
| 5500 |
+
"rewards/rejected": -21.0,
|
| 5501 |
+
"step": 3630
|
| 5502 |
+
},
|
| 5503 |
+
{
|
| 5504 |
+
"epoch": 2.9166666666666665,
|
| 5505 |
+
"grad_norm": 2.0820327560442435e-05,
|
| 5506 |
+
"learning_rate": 1.5434847135648558e-08,
|
| 5507 |
+
"logits/chosen": 1.0,
|
| 5508 |
+
"logits/rejected": 1.4140625,
|
| 5509 |
+
"logps/chosen": -125.0,
|
| 5510 |
+
"logps/rejected": -336.0,
|
| 5511 |
+
"loss": 0.0,
|
| 5512 |
+
"rewards/accuracies": 1.0,
|
| 5513 |
+
"rewards/chosen": 3.078125,
|
| 5514 |
+
"rewards/margins": 23.625,
|
| 5515 |
+
"rewards/rejected": -20.625,
|
| 5516 |
+
"step": 3640
|
| 5517 |
+
},
|
| 5518 |
+
{
|
| 5519 |
+
"epoch": 2.9246794871794872,
|
| 5520 |
+
"grad_norm": 9.867503905955064e-07,
|
| 5521 |
+
"learning_rate": 1.3950727218759274e-08,
|
| 5522 |
+
"logits/chosen": 0.8359375,
|
| 5523 |
+
"logits/rejected": 1.2421875,
|
| 5524 |
+
"logps/chosen": -178.0,
|
| 5525 |
+
"logps/rejected": -354.0,
|
| 5526 |
+
"loss": 0.0,
|
| 5527 |
+
"rewards/accuracies": 1.0,
|
| 5528 |
+
"rewards/chosen": 3.328125,
|
| 5529 |
+
"rewards/margins": 24.75,
|
| 5530 |
+
"rewards/rejected": -21.5,
|
| 5531 |
+
"step": 3650
|
| 5532 |
+
},
|
| 5533 |
+
{
|
| 5534 |
+
"epoch": 2.9326923076923075,
|
| 5535 |
+
"grad_norm": 3.192135749286814e-05,
|
| 5536 |
+
"learning_rate": 1.2466607301869991e-08,
|
| 5537 |
+
"logits/chosen": 0.88671875,
|
| 5538 |
+
"logits/rejected": 1.3671875,
|
| 5539 |
+
"logps/chosen": -144.0,
|
| 5540 |
+
"logps/rejected": -348.0,
|
| 5541 |
+
"loss": 0.0,
|
| 5542 |
+
"rewards/accuracies": 1.0,
|
| 5543 |
+
"rewards/chosen": 3.578125,
|
| 5544 |
+
"rewards/margins": 24.875,
|
| 5545 |
+
"rewards/rejected": -21.25,
|
| 5546 |
+
"step": 3660
|
| 5547 |
+
},
|
| 5548 |
+
{
|
| 5549 |
+
"epoch": 2.940705128205128,
|
| 5550 |
+
"grad_norm": 2.715406001289612e-08,
|
| 5551 |
+
"learning_rate": 1.0982487384980706e-08,
|
| 5552 |
+
"logits/chosen": 1.046875,
|
| 5553 |
+
"logits/rejected": 1.203125,
|
| 5554 |
+
"logps/chosen": -157.0,
|
| 5555 |
+
"logps/rejected": -344.0,
|
| 5556 |
+
"loss": 0.0,
|
| 5557 |
+
"rewards/accuracies": 1.0,
|
| 5558 |
+
"rewards/chosen": 3.4375,
|
| 5559 |
+
"rewards/margins": 24.125,
|
| 5560 |
+
"rewards/rejected": -20.75,
|
| 5561 |
+
"step": 3670
|
| 5562 |
+
},
|
| 5563 |
+
{
|
| 5564 |
+
"epoch": 2.948717948717949,
|
| 5565 |
+
"grad_norm": 2.672696571203794e-07,
|
| 5566 |
+
"learning_rate": 9.498367468091422e-09,
|
| 5567 |
+
"logits/chosen": 0.78515625,
|
| 5568 |
+
"logits/rejected": 1.1484375,
|
| 5569 |
+
"logps/chosen": -177.0,
|
| 5570 |
+
"logps/rejected": -348.0,
|
| 5571 |
+
"loss": 0.0,
|
| 5572 |
+
"rewards/accuracies": 1.0,
|
| 5573 |
+
"rewards/chosen": 3.890625,
|
| 5574 |
+
"rewards/margins": 24.875,
|
| 5575 |
+
"rewards/rejected": -21.0,
|
| 5576 |
+
"step": 3680
|
| 5577 |
+
},
|
| 5578 |
+
{
|
| 5579 |
+
"epoch": 2.956730769230769,
|
| 5580 |
+
"grad_norm": 7.489910666014967e-05,
|
| 5581 |
+
"learning_rate": 8.014247551202137e-09,
|
| 5582 |
+
"logits/chosen": 0.54296875,
|
| 5583 |
+
"logits/rejected": 1.1953125,
|
| 5584 |
+
"logps/chosen": -206.0,
|
| 5585 |
+
"logps/rejected": -338.0,
|
| 5586 |
+
"loss": 0.0,
|
| 5587 |
+
"rewards/accuracies": 1.0,
|
| 5588 |
+
"rewards/chosen": 3.75,
|
| 5589 |
+
"rewards/margins": 23.625,
|
| 5590 |
+
"rewards/rejected": -19.875,
|
| 5591 |
+
"step": 3690
|
| 5592 |
+
},
|
| 5593 |
+
{
|
| 5594 |
+
"epoch": 2.96474358974359,
|
| 5595 |
+
"grad_norm": 1.3844115894190923e-05,
|
| 5596 |
+
"learning_rate": 6.530127634312852e-09,
|
| 5597 |
+
"logits/chosen": 0.85546875,
|
| 5598 |
+
"logits/rejected": 1.0703125,
|
| 5599 |
+
"logps/chosen": -172.0,
|
| 5600 |
+
"logps/rejected": -328.0,
|
| 5601 |
+
"loss": 0.0,
|
| 5602 |
+
"rewards/accuracies": 1.0,
|
| 5603 |
+
"rewards/chosen": 3.671875,
|
| 5604 |
+
"rewards/margins": 22.5,
|
| 5605 |
+
"rewards/rejected": -18.875,
|
| 5606 |
+
"step": 3700
|
| 5607 |
+
},
|
| 5608 |
+
{
|
| 5609 |
+
"epoch": 2.97275641025641,
|
| 5610 |
+
"grad_norm": 2.3143826423249172e-05,
|
| 5611 |
+
"learning_rate": 5.046007717423567e-09,
|
| 5612 |
+
"logits/chosen": 1.0390625,
|
| 5613 |
+
"logits/rejected": 1.0390625,
|
| 5614 |
+
"logps/chosen": -152.0,
|
| 5615 |
+
"logps/rejected": -332.0,
|
| 5616 |
+
"loss": 0.0,
|
| 5617 |
+
"rewards/accuracies": 1.0,
|
| 5618 |
+
"rewards/chosen": 3.46875,
|
| 5619 |
+
"rewards/margins": 24.75,
|
| 5620 |
+
"rewards/rejected": -21.25,
|
| 5621 |
+
"step": 3710
|
| 5622 |
+
},
|
| 5623 |
+
{
|
| 5624 |
+
"epoch": 2.980769230769231,
|
| 5625 |
+
"grad_norm": 9.095521401560934e-05,
|
| 5626 |
+
"learning_rate": 3.561887800534283e-09,
|
| 5627 |
+
"logits/chosen": 0.90625,
|
| 5628 |
+
"logits/rejected": 0.86328125,
|
| 5629 |
+
"logps/chosen": -127.0,
|
| 5630 |
+
"logps/rejected": -326.0,
|
| 5631 |
+
"loss": 0.0002,
|
| 5632 |
+
"rewards/accuracies": 1.0,
|
| 5633 |
+
"rewards/chosen": 3.171875,
|
| 5634 |
+
"rewards/margins": 23.75,
|
| 5635 |
+
"rewards/rejected": -20.5,
|
| 5636 |
+
"step": 3720
|
| 5637 |
+
},
|
| 5638 |
+
{
|
| 5639 |
+
"epoch": 2.988782051282051,
|
| 5640 |
+
"grad_norm": 2.7589599160714447e-07,
|
| 5641 |
+
"learning_rate": 2.0777678836449987e-09,
|
| 5642 |
+
"logits/chosen": 0.73046875,
|
| 5643 |
+
"logits/rejected": 1.078125,
|
| 5644 |
+
"logps/chosen": -193.0,
|
| 5645 |
+
"logps/rejected": -336.0,
|
| 5646 |
+
"loss": 0.0,
|
| 5647 |
+
"rewards/accuracies": 1.0,
|
| 5648 |
+
"rewards/chosen": 3.25,
|
| 5649 |
+
"rewards/margins": 24.5,
|
| 5650 |
+
"rewards/rejected": -21.125,
|
| 5651 |
+
"step": 3730
|
| 5652 |
+
},
|
| 5653 |
+
{
|
| 5654 |
+
"epoch": 2.996794871794872,
|
| 5655 |
+
"grad_norm": 5.964151597883472e-05,
|
| 5656 |
+
"learning_rate": 5.936479667557139e-10,
|
| 5657 |
+
"logits/chosen": 0.83203125,
|
| 5658 |
+
"logits/rejected": 1.3046875,
|
| 5659 |
+
"logps/chosen": -153.0,
|
| 5660 |
+
"logps/rejected": -336.0,
|
| 5661 |
+
"loss": 0.0,
|
| 5662 |
+
"rewards/accuracies": 1.0,
|
| 5663 |
+
"rewards/chosen": 3.609375,
|
| 5664 |
+
"rewards/margins": 23.125,
|
| 5665 |
+
"rewards/rejected": -19.5,
|
| 5666 |
+
"step": 3740
|
| 5667 |
+
},
|
| 5668 |
+
{
|
| 5669 |
+
"epoch": 3.0,
|
| 5670 |
+
"eval_logits/chosen": 0.88671875,
|
| 5671 |
+
"eval_logits/rejected": 1.4453125,
|
| 5672 |
+
"eval_logps/chosen": -170.0,
|
| 5673 |
+
"eval_logps/rejected": -344.0,
|
| 5674 |
+
"eval_loss": 2.2770025225327117e-06,
|
| 5675 |
+
"eval_rewards/accuracies": 1.0,
|
| 5676 |
+
"eval_rewards/chosen": 3.296875,
|
| 5677 |
+
"eval_rewards/margins": 23.5,
|
| 5678 |
+
"eval_rewards/rejected": -20.25,
|
| 5679 |
+
"eval_runtime": 32.9212,
|
| 5680 |
+
"eval_samples_per_second": 6.045,
|
| 5681 |
+
"eval_steps_per_second": 0.759,
|
| 5682 |
+
"step": 3744
|
| 5683 |
}
|
| 5684 |
],
|
| 5685 |
"logging_steps": 10,
|
|
|
|
| 5694 |
"should_evaluate": false,
|
| 5695 |
"should_log": false,
|
| 5696 |
"should_save": true,
|
| 5697 |
+
"should_training_stop": true
|
| 5698 |
},
|
| 5699 |
"attributes": {}
|
| 5700 |
}
|