Training in progress, epoch 2, checkpoint
Browse files- last-checkpoint/global_step1248/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step1248/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step1248/zero_pp_rank_0_mp_rank_00_model_states.pt +3 -0
- last-checkpoint/global_step1248/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 +948 -2
last-checkpoint/global_step1248/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:c31ae94d6da4b033fe866a1018178a1f7ae5a72f6200ed71c44febd9a0c24ae6
|
| 3 |
+
size 30462473157
|
last-checkpoint/global_step1248/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:fbd331d7c4e7eba50f28bb9f427bbba4828992a5e8f5a691e7f93803cb74bf3b
|
| 3 |
+
size 30462473157
|
last-checkpoint/global_step1248/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 sha256:5d1718af0fe3e085976bfada2ffb44c751c2d06c64ed7cc32035b9588c36b290
|
| 3 |
+
size 168021
|
last-checkpoint/global_step1248/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:f3bb56930600447fb01156386fcad548b765911ec81559d7b193afbcb58d6930
|
| 3 |
+
size 168021
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step1248
|
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 sha256:30dd02bdd9a1699ec1b6185608050346737c71cf556d27d7a82e5d9ce434fdd0
|
| 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 sha256:025ca15fd6ee18e66623c43e1b6715ec460b3c1db4b5ae83059f58b01f690c9f
|
| 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 sha256:584a552a41e564fcd7f833dc0769480f6c7fd9cd9d71a3fe96f234305784179b
|
| 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 sha256:6b7aff33fb6e1e771d13945f4cb5b44d4da73a682d5a6e9366a6386df2908523
|
| 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/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:b3b7bf841d31595f02eb1941a019ed4bf66f16ec16e6c5eb963a853209da5c2f
|
| 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,
|
|
@@ -968,6 +968,952 @@
|
|
| 968 |
"eval_samples_per_second": 7.126,
|
| 969 |
"eval_steps_per_second": 0.463,
|
| 970 |
"step": 624
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 971 |
}
|
| 972 |
],
|
| 973 |
"logging_steps": 10,
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 1248,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 968 |
"eval_samples_per_second": 7.126,
|
| 969 |
"eval_steps_per_second": 0.463,
|
| 970 |
"step": 624
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.0096153846153846,
|
| 974 |
+
"grad_norm": 6.092991588272169,
|
| 975 |
+
"learning_rate": 3.687648456057007e-07,
|
| 976 |
+
"logits/chosen": -0.2734375,
|
| 977 |
+
"logits/rejected": -0.388671875,
|
| 978 |
+
"logps/chosen": -350.0,
|
| 979 |
+
"logps/rejected": -322.0,
|
| 980 |
+
"loss": 0.1514,
|
| 981 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 982 |
+
"rewards/chosen": 1.7734375,
|
| 983 |
+
"rewards/margins": 3.96875,
|
| 984 |
+
"rewards/rejected": -2.203125,
|
| 985 |
+
"step": 630
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 1.0256410256410255,
|
| 989 |
+
"grad_norm": 15.11610385237531,
|
| 990 |
+
"learning_rate": 3.6579572446555817e-07,
|
| 991 |
+
"logits/chosen": -0.21875,
|
| 992 |
+
"logits/rejected": -0.46484375,
|
| 993 |
+
"logps/chosen": -356.0,
|
| 994 |
+
"logps/rejected": -374.0,
|
| 995 |
+
"loss": 0.0783,
|
| 996 |
+
"rewards/accuracies": 0.987500011920929,
|
| 997 |
+
"rewards/chosen": 1.921875,
|
| 998 |
+
"rewards/margins": 4.8125,
|
| 999 |
+
"rewards/rejected": -2.890625,
|
| 1000 |
+
"step": 640
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 1.0416666666666667,
|
| 1004 |
+
"grad_norm": 8.924081971380067,
|
| 1005 |
+
"learning_rate": 3.6282660332541565e-07,
|
| 1006 |
+
"logits/chosen": -0.36328125,
|
| 1007 |
+
"logits/rejected": -0.5,
|
| 1008 |
+
"logps/chosen": -378.0,
|
| 1009 |
+
"logps/rejected": -348.0,
|
| 1010 |
+
"loss": 0.0486,
|
| 1011 |
+
"rewards/accuracies": 1.0,
|
| 1012 |
+
"rewards/chosen": 1.7421875,
|
| 1013 |
+
"rewards/margins": 4.59375,
|
| 1014 |
+
"rewards/rejected": -2.84375,
|
| 1015 |
+
"step": 650
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"epoch": 1.0576923076923077,
|
| 1019 |
+
"grad_norm": 8.969543433658254,
|
| 1020 |
+
"learning_rate": 3.598574821852731e-07,
|
| 1021 |
+
"logits/chosen": -0.1572265625,
|
| 1022 |
+
"logits/rejected": -0.2890625,
|
| 1023 |
+
"logps/chosen": -364.0,
|
| 1024 |
+
"logps/rejected": -344.0,
|
| 1025 |
+
"loss": 0.1064,
|
| 1026 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1027 |
+
"rewards/chosen": 2.109375,
|
| 1028 |
+
"rewards/margins": 4.1875,
|
| 1029 |
+
"rewards/rejected": -2.078125,
|
| 1030 |
+
"step": 660
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 1.0737179487179487,
|
| 1034 |
+
"grad_norm": 20.194350083453877,
|
| 1035 |
+
"learning_rate": 3.5688836104513066e-07,
|
| 1036 |
+
"logits/chosen": -0.1884765625,
|
| 1037 |
+
"logits/rejected": -0.427734375,
|
| 1038 |
+
"logps/chosen": -376.0,
|
| 1039 |
+
"logps/rejected": -326.0,
|
| 1040 |
+
"loss": 0.0802,
|
| 1041 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1042 |
+
"rewards/chosen": 2.109375,
|
| 1043 |
+
"rewards/margins": 4.3125,
|
| 1044 |
+
"rewards/rejected": -2.1875,
|
| 1045 |
+
"step": 670
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.0897435897435896,
|
| 1049 |
+
"grad_norm": 9.472914566833554,
|
| 1050 |
+
"learning_rate": 3.5391923990498813e-07,
|
| 1051 |
+
"logits/chosen": -0.1494140625,
|
| 1052 |
+
"logits/rejected": -0.39453125,
|
| 1053 |
+
"logps/chosen": -354.0,
|
| 1054 |
+
"logps/rejected": -330.0,
|
| 1055 |
+
"loss": 0.0748,
|
| 1056 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1057 |
+
"rewards/chosen": 2.671875,
|
| 1058 |
+
"rewards/margins": 5.0625,
|
| 1059 |
+
"rewards/rejected": -2.390625,
|
| 1060 |
+
"step": 680
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.1057692307692308,
|
| 1064 |
+
"grad_norm": 27.091485921309573,
|
| 1065 |
+
"learning_rate": 3.509501187648456e-07,
|
| 1066 |
+
"logits/chosen": -0.2578125,
|
| 1067 |
+
"logits/rejected": -0.359375,
|
| 1068 |
+
"logps/chosen": -368.0,
|
| 1069 |
+
"logps/rejected": -324.0,
|
| 1070 |
+
"loss": 0.0862,
|
| 1071 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1072 |
+
"rewards/chosen": 2.140625,
|
| 1073 |
+
"rewards/margins": 4.71875,
|
| 1074 |
+
"rewards/rejected": -2.5625,
|
| 1075 |
+
"step": 690
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.1217948717948718,
|
| 1079 |
+
"grad_norm": 23.55129242956029,
|
| 1080 |
+
"learning_rate": 3.479809976247031e-07,
|
| 1081 |
+
"logits/chosen": -0.1875,
|
| 1082 |
+
"logits/rejected": -0.25,
|
| 1083 |
+
"logps/chosen": -352.0,
|
| 1084 |
+
"logps/rejected": -314.0,
|
| 1085 |
+
"loss": 0.0667,
|
| 1086 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1087 |
+
"rewards/chosen": 2.21875,
|
| 1088 |
+
"rewards/margins": 4.65625,
|
| 1089 |
+
"rewards/rejected": -2.4375,
|
| 1090 |
+
"step": 700
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 1.1378205128205128,
|
| 1094 |
+
"grad_norm": 16.191916223045457,
|
| 1095 |
+
"learning_rate": 3.450118764845605e-07,
|
| 1096 |
+
"logits/chosen": -0.2431640625,
|
| 1097 |
+
"logits/rejected": -0.4453125,
|
| 1098 |
+
"logps/chosen": -368.0,
|
| 1099 |
+
"logps/rejected": -352.0,
|
| 1100 |
+
"loss": 0.1058,
|
| 1101 |
+
"rewards/accuracies": 0.9375,
|
| 1102 |
+
"rewards/chosen": 1.953125,
|
| 1103 |
+
"rewards/margins": 4.84375,
|
| 1104 |
+
"rewards/rejected": -2.890625,
|
| 1105 |
+
"step": 710
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 1.1538461538461537,
|
| 1109 |
+
"grad_norm": 6.197082349513418,
|
| 1110 |
+
"learning_rate": 3.42042755344418e-07,
|
| 1111 |
+
"logits/chosen": -0.193359375,
|
| 1112 |
+
"logits/rejected": -0.298828125,
|
| 1113 |
+
"logps/chosen": -420.0,
|
| 1114 |
+
"logps/rejected": -352.0,
|
| 1115 |
+
"loss": 0.1152,
|
| 1116 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1117 |
+
"rewards/chosen": 1.140625,
|
| 1118 |
+
"rewards/margins": 4.8125,
|
| 1119 |
+
"rewards/rejected": -3.65625,
|
| 1120 |
+
"step": 720
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 1.169871794871795,
|
| 1124 |
+
"grad_norm": 20.175263771423996,
|
| 1125 |
+
"learning_rate": 3.390736342042755e-07,
|
| 1126 |
+
"logits/chosen": -0.0849609375,
|
| 1127 |
+
"logits/rejected": -0.30859375,
|
| 1128 |
+
"logps/chosen": -368.0,
|
| 1129 |
+
"logps/rejected": -340.0,
|
| 1130 |
+
"loss": 0.0717,
|
| 1131 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1132 |
+
"rewards/chosen": 1.8203125,
|
| 1133 |
+
"rewards/margins": 5.125,
|
| 1134 |
+
"rewards/rejected": -3.296875,
|
| 1135 |
+
"step": 730
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 1.185897435897436,
|
| 1139 |
+
"grad_norm": 14.053624440757009,
|
| 1140 |
+
"learning_rate": 3.36104513064133e-07,
|
| 1141 |
+
"logits/chosen": -0.1591796875,
|
| 1142 |
+
"logits/rejected": -0.373046875,
|
| 1143 |
+
"logps/chosen": -368.0,
|
| 1144 |
+
"logps/rejected": -342.0,
|
| 1145 |
+
"loss": 0.0763,
|
| 1146 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1147 |
+
"rewards/chosen": 2.0,
|
| 1148 |
+
"rewards/margins": 5.53125,
|
| 1149 |
+
"rewards/rejected": -3.515625,
|
| 1150 |
+
"step": 740
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.2019230769230769,
|
| 1154 |
+
"grad_norm": 19.07797922710506,
|
| 1155 |
+
"learning_rate": 3.331353919239905e-07,
|
| 1156 |
+
"logits/chosen": -0.134765625,
|
| 1157 |
+
"logits/rejected": -0.37109375,
|
| 1158 |
+
"logps/chosen": -374.0,
|
| 1159 |
+
"logps/rejected": -346.0,
|
| 1160 |
+
"loss": 0.0837,
|
| 1161 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1162 |
+
"rewards/chosen": 1.859375,
|
| 1163 |
+
"rewards/margins": 4.9375,
|
| 1164 |
+
"rewards/rejected": -3.078125,
|
| 1165 |
+
"step": 750
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 1.217948717948718,
|
| 1169 |
+
"grad_norm": 3.621341208921815,
|
| 1170 |
+
"learning_rate": 3.3016627078384796e-07,
|
| 1171 |
+
"logits/chosen": -0.1337890625,
|
| 1172 |
+
"logits/rejected": -0.447265625,
|
| 1173 |
+
"logps/chosen": -382.0,
|
| 1174 |
+
"logps/rejected": -362.0,
|
| 1175 |
+
"loss": 0.0574,
|
| 1176 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1177 |
+
"rewards/chosen": 1.7734375,
|
| 1178 |
+
"rewards/margins": 5.0625,
|
| 1179 |
+
"rewards/rejected": -3.28125,
|
| 1180 |
+
"step": 760
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 1.233974358974359,
|
| 1184 |
+
"grad_norm": 32.55125289195706,
|
| 1185 |
+
"learning_rate": 3.2719714964370544e-07,
|
| 1186 |
+
"logits/chosen": -0.2099609375,
|
| 1187 |
+
"logits/rejected": -0.408203125,
|
| 1188 |
+
"logps/chosen": -354.0,
|
| 1189 |
+
"logps/rejected": -330.0,
|
| 1190 |
+
"loss": 0.0801,
|
| 1191 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1192 |
+
"rewards/chosen": 2.203125,
|
| 1193 |
+
"rewards/margins": 5.6875,
|
| 1194 |
+
"rewards/rejected": -3.46875,
|
| 1195 |
+
"step": 770
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 1.25,
|
| 1199 |
+
"grad_norm": 22.82618457101383,
|
| 1200 |
+
"learning_rate": 3.2422802850356297e-07,
|
| 1201 |
+
"logits/chosen": -0.265625,
|
| 1202 |
+
"logits/rejected": -0.486328125,
|
| 1203 |
+
"logps/chosen": -362.0,
|
| 1204 |
+
"logps/rejected": -334.0,
|
| 1205 |
+
"loss": 0.0937,
|
| 1206 |
+
"rewards/accuracies": 0.925000011920929,
|
| 1207 |
+
"rewards/chosen": 2.03125,
|
| 1208 |
+
"rewards/margins": 5.0625,
|
| 1209 |
+
"rewards/rejected": -3.046875,
|
| 1210 |
+
"step": 780
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 1.266025641025641,
|
| 1214 |
+
"grad_norm": 49.835975518914886,
|
| 1215 |
+
"learning_rate": 3.2125890736342045e-07,
|
| 1216 |
+
"logits/chosen": -0.24609375,
|
| 1217 |
+
"logits/rejected": -0.447265625,
|
| 1218 |
+
"logps/chosen": -360.0,
|
| 1219 |
+
"logps/rejected": -352.0,
|
| 1220 |
+
"loss": 0.0787,
|
| 1221 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1222 |
+
"rewards/chosen": 2.140625,
|
| 1223 |
+
"rewards/margins": 5.5,
|
| 1224 |
+
"rewards/rejected": -3.375,
|
| 1225 |
+
"step": 790
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 1.282051282051282,
|
| 1229 |
+
"grad_norm": 3.1313003868772107,
|
| 1230 |
+
"learning_rate": 3.182897862232779e-07,
|
| 1231 |
+
"logits/chosen": -0.283203125,
|
| 1232 |
+
"logits/rejected": -0.287109375,
|
| 1233 |
+
"logps/chosen": -372.0,
|
| 1234 |
+
"logps/rejected": -326.0,
|
| 1235 |
+
"loss": 0.1086,
|
| 1236 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1237 |
+
"rewards/chosen": 1.6171875,
|
| 1238 |
+
"rewards/margins": 4.78125,
|
| 1239 |
+
"rewards/rejected": -3.1875,
|
| 1240 |
+
"step": 800
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 1.2980769230769231,
|
| 1244 |
+
"grad_norm": 13.950684325273025,
|
| 1245 |
+
"learning_rate": 3.1532066508313535e-07,
|
| 1246 |
+
"logits/chosen": -0.234375,
|
| 1247 |
+
"logits/rejected": -0.640625,
|
| 1248 |
+
"logps/chosen": -344.0,
|
| 1249 |
+
"logps/rejected": -344.0,
|
| 1250 |
+
"loss": 0.0843,
|
| 1251 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1252 |
+
"rewards/chosen": 2.140625,
|
| 1253 |
+
"rewards/margins": 5.0625,
|
| 1254 |
+
"rewards/rejected": -2.90625,
|
| 1255 |
+
"step": 810
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.314102564102564,
|
| 1259 |
+
"grad_norm": 6.930473305635045,
|
| 1260 |
+
"learning_rate": 3.1235154394299283e-07,
|
| 1261 |
+
"logits/chosen": -0.341796875,
|
| 1262 |
+
"logits/rejected": -0.326171875,
|
| 1263 |
+
"logps/chosen": -362.0,
|
| 1264 |
+
"logps/rejected": -344.0,
|
| 1265 |
+
"loss": 0.0667,
|
| 1266 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1267 |
+
"rewards/chosen": 1.984375,
|
| 1268 |
+
"rewards/margins": 5.21875,
|
| 1269 |
+
"rewards/rejected": -3.234375,
|
| 1270 |
+
"step": 820
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 1.330128205128205,
|
| 1274 |
+
"grad_norm": 21.31500643130334,
|
| 1275 |
+
"learning_rate": 3.093824228028503e-07,
|
| 1276 |
+
"logits/chosen": -0.1748046875,
|
| 1277 |
+
"logits/rejected": -0.443359375,
|
| 1278 |
+
"logps/chosen": -366.0,
|
| 1279 |
+
"logps/rejected": -342.0,
|
| 1280 |
+
"loss": 0.0595,
|
| 1281 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1282 |
+
"rewards/chosen": 1.984375,
|
| 1283 |
+
"rewards/margins": 5.5625,
|
| 1284 |
+
"rewards/rejected": -3.5625,
|
| 1285 |
+
"step": 830
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 1.3461538461538463,
|
| 1289 |
+
"grad_norm": 32.89803571469975,
|
| 1290 |
+
"learning_rate": 3.0641330166270784e-07,
|
| 1291 |
+
"logits/chosen": -0.234375,
|
| 1292 |
+
"logits/rejected": -0.34375,
|
| 1293 |
+
"logps/chosen": -330.0,
|
| 1294 |
+
"logps/rejected": -374.0,
|
| 1295 |
+
"loss": 0.0511,
|
| 1296 |
+
"rewards/accuracies": 1.0,
|
| 1297 |
+
"rewards/chosen": 2.28125,
|
| 1298 |
+
"rewards/margins": 6.0625,
|
| 1299 |
+
"rewards/rejected": -3.8125,
|
| 1300 |
+
"step": 840
|
| 1301 |
+
},
|
| 1302 |
+
{
|
| 1303 |
+
"epoch": 1.3621794871794872,
|
| 1304 |
+
"grad_norm": 32.22922106524605,
|
| 1305 |
+
"learning_rate": 3.034441805225653e-07,
|
| 1306 |
+
"logits/chosen": -0.279296875,
|
| 1307 |
+
"logits/rejected": -0.32421875,
|
| 1308 |
+
"logps/chosen": -358.0,
|
| 1309 |
+
"logps/rejected": -358.0,
|
| 1310 |
+
"loss": 0.0406,
|
| 1311 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1312 |
+
"rewards/chosen": 2.609375,
|
| 1313 |
+
"rewards/margins": 6.21875,
|
| 1314 |
+
"rewards/rejected": -3.59375,
|
| 1315 |
+
"step": 850
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 1.3782051282051282,
|
| 1319 |
+
"grad_norm": 28.46125519366018,
|
| 1320 |
+
"learning_rate": 3.004750593824228e-07,
|
| 1321 |
+
"logits/chosen": -0.2294921875,
|
| 1322 |
+
"logits/rejected": -0.19140625,
|
| 1323 |
+
"logps/chosen": -348.0,
|
| 1324 |
+
"logps/rejected": -330.0,
|
| 1325 |
+
"loss": 0.0549,
|
| 1326 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1327 |
+
"rewards/chosen": 2.109375,
|
| 1328 |
+
"rewards/margins": 6.15625,
|
| 1329 |
+
"rewards/rejected": -4.0625,
|
| 1330 |
+
"step": 860
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 1.3942307692307692,
|
| 1334 |
+
"grad_norm": 14.047872703813512,
|
| 1335 |
+
"learning_rate": 2.975059382422803e-07,
|
| 1336 |
+
"logits/chosen": -0.2490234375,
|
| 1337 |
+
"logits/rejected": -0.359375,
|
| 1338 |
+
"logps/chosen": -366.0,
|
| 1339 |
+
"logps/rejected": -316.0,
|
| 1340 |
+
"loss": 0.0797,
|
| 1341 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1342 |
+
"rewards/chosen": 2.296875,
|
| 1343 |
+
"rewards/margins": 6.375,
|
| 1344 |
+
"rewards/rejected": -4.09375,
|
| 1345 |
+
"step": 870
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 1.4102564102564101,
|
| 1349 |
+
"grad_norm": 11.52369059409131,
|
| 1350 |
+
"learning_rate": 2.9453681710213776e-07,
|
| 1351 |
+
"logits/chosen": -0.26953125,
|
| 1352 |
+
"logits/rejected": -0.359375,
|
| 1353 |
+
"logps/chosen": -400.0,
|
| 1354 |
+
"logps/rejected": -348.0,
|
| 1355 |
+
"loss": 0.052,
|
| 1356 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1357 |
+
"rewards/chosen": 2.046875,
|
| 1358 |
+
"rewards/margins": 5.90625,
|
| 1359 |
+
"rewards/rejected": -3.84375,
|
| 1360 |
+
"step": 880
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.4262820512820513,
|
| 1364 |
+
"grad_norm": 15.524036946329323,
|
| 1365 |
+
"learning_rate": 2.915676959619953e-07,
|
| 1366 |
+
"logits/chosen": -0.1923828125,
|
| 1367 |
+
"logits/rejected": -0.283203125,
|
| 1368 |
+
"logps/chosen": -364.0,
|
| 1369 |
+
"logps/rejected": -358.0,
|
| 1370 |
+
"loss": 0.0819,
|
| 1371 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1372 |
+
"rewards/chosen": 2.125,
|
| 1373 |
+
"rewards/margins": 6.1875,
|
| 1374 |
+
"rewards/rejected": -4.0625,
|
| 1375 |
+
"step": 890
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 1.4423076923076923,
|
| 1379 |
+
"grad_norm": 56.41338438001929,
|
| 1380 |
+
"learning_rate": 2.885985748218527e-07,
|
| 1381 |
+
"logits/chosen": -0.2314453125,
|
| 1382 |
+
"logits/rejected": -0.427734375,
|
| 1383 |
+
"logps/chosen": -374.0,
|
| 1384 |
+
"logps/rejected": -352.0,
|
| 1385 |
+
"loss": 0.1086,
|
| 1386 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1387 |
+
"rewards/chosen": 1.546875,
|
| 1388 |
+
"rewards/margins": 5.65625,
|
| 1389 |
+
"rewards/rejected": -4.125,
|
| 1390 |
+
"step": 900
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 1.4583333333333333,
|
| 1394 |
+
"grad_norm": 35.802404269657636,
|
| 1395 |
+
"learning_rate": 2.856294536817102e-07,
|
| 1396 |
+
"logits/chosen": -0.169921875,
|
| 1397 |
+
"logits/rejected": -0.296875,
|
| 1398 |
+
"logps/chosen": -366.0,
|
| 1399 |
+
"logps/rejected": -352.0,
|
| 1400 |
+
"loss": 0.0569,
|
| 1401 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1402 |
+
"rewards/chosen": 2.09375,
|
| 1403 |
+
"rewards/margins": 5.8125,
|
| 1404 |
+
"rewards/rejected": -3.71875,
|
| 1405 |
+
"step": 910
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 1.4743589743589745,
|
| 1409 |
+
"grad_norm": 3.0990206807529574,
|
| 1410 |
+
"learning_rate": 2.8266033254156767e-07,
|
| 1411 |
+
"logits/chosen": -0.310546875,
|
| 1412 |
+
"logits/rejected": -0.16015625,
|
| 1413 |
+
"logps/chosen": -364.0,
|
| 1414 |
+
"logps/rejected": -358.0,
|
| 1415 |
+
"loss": 0.0532,
|
| 1416 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1417 |
+
"rewards/chosen": 2.359375,
|
| 1418 |
+
"rewards/margins": 5.75,
|
| 1419 |
+
"rewards/rejected": -3.390625,
|
| 1420 |
+
"step": 920
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 1.4903846153846154,
|
| 1424 |
+
"grad_norm": 6.630507067800798,
|
| 1425 |
+
"learning_rate": 2.7969121140142515e-07,
|
| 1426 |
+
"logits/chosen": -0.2099609375,
|
| 1427 |
+
"logits/rejected": -0.384765625,
|
| 1428 |
+
"logps/chosen": -352.0,
|
| 1429 |
+
"logps/rejected": -344.0,
|
| 1430 |
+
"loss": 0.1117,
|
| 1431 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1432 |
+
"rewards/chosen": 2.78125,
|
| 1433 |
+
"rewards/margins": 5.9375,
|
| 1434 |
+
"rewards/rejected": -3.140625,
|
| 1435 |
+
"step": 930
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 1.5064102564102564,
|
| 1439 |
+
"grad_norm": 3.9527473386787992,
|
| 1440 |
+
"learning_rate": 2.7672209026128263e-07,
|
| 1441 |
+
"logits/chosen": -0.150390625,
|
| 1442 |
+
"logits/rejected": -0.353515625,
|
| 1443 |
+
"logps/chosen": -364.0,
|
| 1444 |
+
"logps/rejected": -326.0,
|
| 1445 |
+
"loss": 0.1115,
|
| 1446 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1447 |
+
"rewards/chosen": 2.71875,
|
| 1448 |
+
"rewards/margins": 5.375,
|
| 1449 |
+
"rewards/rejected": -2.65625,
|
| 1450 |
+
"step": 940
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 1.5224358974358974,
|
| 1454 |
+
"grad_norm": 21.74505432942635,
|
| 1455 |
+
"learning_rate": 2.7375296912114016e-07,
|
| 1456 |
+
"logits/chosen": -0.224609375,
|
| 1457 |
+
"logits/rejected": -0.3203125,
|
| 1458 |
+
"logps/chosen": -388.0,
|
| 1459 |
+
"logps/rejected": -364.0,
|
| 1460 |
+
"loss": 0.0695,
|
| 1461 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1462 |
+
"rewards/chosen": 2.765625,
|
| 1463 |
+
"rewards/margins": 5.5,
|
| 1464 |
+
"rewards/rejected": -2.75,
|
| 1465 |
+
"step": 950
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 1.5384615384615383,
|
| 1469 |
+
"grad_norm": 34.5728416770996,
|
| 1470 |
+
"learning_rate": 2.7078384798099764e-07,
|
| 1471 |
+
"logits/chosen": -0.2197265625,
|
| 1472 |
+
"logits/rejected": -0.412109375,
|
| 1473 |
+
"logps/chosen": -356.0,
|
| 1474 |
+
"logps/rejected": -328.0,
|
| 1475 |
+
"loss": 0.0797,
|
| 1476 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1477 |
+
"rewards/chosen": 2.9375,
|
| 1478 |
+
"rewards/margins": 6.40625,
|
| 1479 |
+
"rewards/rejected": -3.46875,
|
| 1480 |
+
"step": 960
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 1.5544871794871795,
|
| 1484 |
+
"grad_norm": 10.226160428385189,
|
| 1485 |
+
"learning_rate": 2.678147268408551e-07,
|
| 1486 |
+
"logits/chosen": -0.1787109375,
|
| 1487 |
+
"logits/rejected": -0.4140625,
|
| 1488 |
+
"logps/chosen": -372.0,
|
| 1489 |
+
"logps/rejected": -330.0,
|
| 1490 |
+
"loss": 0.0415,
|
| 1491 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1492 |
+
"rewards/chosen": 2.40625,
|
| 1493 |
+
"rewards/margins": 6.40625,
|
| 1494 |
+
"rewards/rejected": -3.984375,
|
| 1495 |
+
"step": 970
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 1.5705128205128205,
|
| 1499 |
+
"grad_norm": 2.887418080486524,
|
| 1500 |
+
"learning_rate": 2.648456057007126e-07,
|
| 1501 |
+
"logits/chosen": -0.1357421875,
|
| 1502 |
+
"logits/rejected": -0.318359375,
|
| 1503 |
+
"logps/chosen": -388.0,
|
| 1504 |
+
"logps/rejected": -320.0,
|
| 1505 |
+
"loss": 0.0622,
|
| 1506 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1507 |
+
"rewards/chosen": 2.65625,
|
| 1508 |
+
"rewards/margins": 6.40625,
|
| 1509 |
+
"rewards/rejected": -3.734375,
|
| 1510 |
+
"step": 980
|
| 1511 |
+
},
|
| 1512 |
+
{
|
| 1513 |
+
"epoch": 1.5865384615384617,
|
| 1514 |
+
"grad_norm": 25.20155218665509,
|
| 1515 |
+
"learning_rate": 2.6187648456057e-07,
|
| 1516 |
+
"logits/chosen": -0.1044921875,
|
| 1517 |
+
"logits/rejected": -0.37890625,
|
| 1518 |
+
"logps/chosen": -330.0,
|
| 1519 |
+
"logps/rejected": -326.0,
|
| 1520 |
+
"loss": 0.0672,
|
| 1521 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1522 |
+
"rewards/chosen": 2.921875,
|
| 1523 |
+
"rewards/margins": 6.75,
|
| 1524 |
+
"rewards/rejected": -3.828125,
|
| 1525 |
+
"step": 990
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 1.6025641025641026,
|
| 1529 |
+
"grad_norm": 2.9326020885487107,
|
| 1530 |
+
"learning_rate": 2.589073634204275e-07,
|
| 1531 |
+
"logits/chosen": -0.244140625,
|
| 1532 |
+
"logits/rejected": -0.4609375,
|
| 1533 |
+
"logps/chosen": -370.0,
|
| 1534 |
+
"logps/rejected": -336.0,
|
| 1535 |
+
"loss": 0.0686,
|
| 1536 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1537 |
+
"rewards/chosen": 2.4375,
|
| 1538 |
+
"rewards/margins": 6.375,
|
| 1539 |
+
"rewards/rejected": -3.9375,
|
| 1540 |
+
"step": 1000
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 1.6185897435897436,
|
| 1544 |
+
"grad_norm": 2.6087260544675948,
|
| 1545 |
+
"learning_rate": 2.5593824228028503e-07,
|
| 1546 |
+
"logits/chosen": -0.330078125,
|
| 1547 |
+
"logits/rejected": -0.59765625,
|
| 1548 |
+
"logps/chosen": -366.0,
|
| 1549 |
+
"logps/rejected": -384.0,
|
| 1550 |
+
"loss": 0.0506,
|
| 1551 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1552 |
+
"rewards/chosen": 2.390625,
|
| 1553 |
+
"rewards/margins": 6.4375,
|
| 1554 |
+
"rewards/rejected": -4.0625,
|
| 1555 |
+
"step": 1010
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 1.6346153846153846,
|
| 1559 |
+
"grad_norm": 46.202092572333505,
|
| 1560 |
+
"learning_rate": 2.529691211401425e-07,
|
| 1561 |
+
"logits/chosen": -0.10791015625,
|
| 1562 |
+
"logits/rejected": -0.40234375,
|
| 1563 |
+
"logps/chosen": -370.0,
|
| 1564 |
+
"logps/rejected": -368.0,
|
| 1565 |
+
"loss": 0.1479,
|
| 1566 |
+
"rewards/accuracies": 0.9375,
|
| 1567 |
+
"rewards/chosen": 1.9609375,
|
| 1568 |
+
"rewards/margins": 5.6875,
|
| 1569 |
+
"rewards/rejected": -3.71875,
|
| 1570 |
+
"step": 1020
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 1.6506410256410255,
|
| 1574 |
+
"grad_norm": 11.039931814429327,
|
| 1575 |
+
"learning_rate": 2.5e-07,
|
| 1576 |
+
"logits/chosen": -0.1875,
|
| 1577 |
+
"logits/rejected": -0.357421875,
|
| 1578 |
+
"logps/chosen": -398.0,
|
| 1579 |
+
"logps/rejected": -348.0,
|
| 1580 |
+
"loss": 0.0695,
|
| 1581 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1582 |
+
"rewards/chosen": 1.6875,
|
| 1583 |
+
"rewards/margins": 5.59375,
|
| 1584 |
+
"rewards/rejected": -3.90625,
|
| 1585 |
+
"step": 1030
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 1.6666666666666665,
|
| 1589 |
+
"grad_norm": 19.29266338720831,
|
| 1590 |
+
"learning_rate": 2.4703087885985747e-07,
|
| 1591 |
+
"logits/chosen": -0.2353515625,
|
| 1592 |
+
"logits/rejected": -0.4609375,
|
| 1593 |
+
"logps/chosen": -364.0,
|
| 1594 |
+
"logps/rejected": -350.0,
|
| 1595 |
+
"loss": 0.0767,
|
| 1596 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1597 |
+
"rewards/chosen": 1.5078125,
|
| 1598 |
+
"rewards/margins": 6.15625,
|
| 1599 |
+
"rewards/rejected": -4.65625,
|
| 1600 |
+
"step": 1040
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 1.6826923076923077,
|
| 1604 |
+
"grad_norm": 42.04473353572841,
|
| 1605 |
+
"learning_rate": 2.4406175771971495e-07,
|
| 1606 |
+
"logits/chosen": -0.375,
|
| 1607 |
+
"logits/rejected": -0.515625,
|
| 1608 |
+
"logps/chosen": -374.0,
|
| 1609 |
+
"logps/rejected": -356.0,
|
| 1610 |
+
"loss": 0.0784,
|
| 1611 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1612 |
+
"rewards/chosen": 1.8046875,
|
| 1613 |
+
"rewards/margins": 5.84375,
|
| 1614 |
+
"rewards/rejected": -4.0625,
|
| 1615 |
+
"step": 1050
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 1.6987179487179487,
|
| 1619 |
+
"grad_norm": 48.05824905529893,
|
| 1620 |
+
"learning_rate": 2.410926365795724e-07,
|
| 1621 |
+
"logits/chosen": -0.2109375,
|
| 1622 |
+
"logits/rejected": -0.353515625,
|
| 1623 |
+
"logps/chosen": -380.0,
|
| 1624 |
+
"logps/rejected": -342.0,
|
| 1625 |
+
"loss": 0.0879,
|
| 1626 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1627 |
+
"rewards/chosen": 1.7109375,
|
| 1628 |
+
"rewards/margins": 5.40625,
|
| 1629 |
+
"rewards/rejected": -3.703125,
|
| 1630 |
+
"step": 1060
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 1.7147435897435899,
|
| 1634 |
+
"grad_norm": 9.491221178753944,
|
| 1635 |
+
"learning_rate": 2.381235154394299e-07,
|
| 1636 |
+
"logits/chosen": -0.2138671875,
|
| 1637 |
+
"logits/rejected": -0.4375,
|
| 1638 |
+
"logps/chosen": -372.0,
|
| 1639 |
+
"logps/rejected": -358.0,
|
| 1640 |
+
"loss": 0.0497,
|
| 1641 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1642 |
+
"rewards/chosen": 2.296875,
|
| 1643 |
+
"rewards/margins": 6.4375,
|
| 1644 |
+
"rewards/rejected": -4.125,
|
| 1645 |
+
"step": 1070
|
| 1646 |
+
},
|
| 1647 |
+
{
|
| 1648 |
+
"epoch": 1.7307692307692308,
|
| 1649 |
+
"grad_norm": 40.31044273857979,
|
| 1650 |
+
"learning_rate": 2.351543942992874e-07,
|
| 1651 |
+
"logits/chosen": -0.248046875,
|
| 1652 |
+
"logits/rejected": -0.51953125,
|
| 1653 |
+
"logps/chosen": -360.0,
|
| 1654 |
+
"logps/rejected": -350.0,
|
| 1655 |
+
"loss": 0.1188,
|
| 1656 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1657 |
+
"rewards/chosen": 2.515625,
|
| 1658 |
+
"rewards/margins": 6.34375,
|
| 1659 |
+
"rewards/rejected": -3.8125,
|
| 1660 |
+
"step": 1080
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 1.7467948717948718,
|
| 1664 |
+
"grad_norm": 1.4266357038353779,
|
| 1665 |
+
"learning_rate": 2.3218527315914489e-07,
|
| 1666 |
+
"logits/chosen": -0.287109375,
|
| 1667 |
+
"logits/rejected": -0.283203125,
|
| 1668 |
+
"logps/chosen": -356.0,
|
| 1669 |
+
"logps/rejected": -328.0,
|
| 1670 |
+
"loss": 0.1075,
|
| 1671 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1672 |
+
"rewards/chosen": 2.265625,
|
| 1673 |
+
"rewards/margins": 5.59375,
|
| 1674 |
+
"rewards/rejected": -3.328125,
|
| 1675 |
+
"step": 1090
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 1.7628205128205128,
|
| 1679 |
+
"grad_norm": 21.16673032403483,
|
| 1680 |
+
"learning_rate": 2.2921615201900234e-07,
|
| 1681 |
+
"logits/chosen": -0.1572265625,
|
| 1682 |
+
"logits/rejected": -0.408203125,
|
| 1683 |
+
"logps/chosen": -392.0,
|
| 1684 |
+
"logps/rejected": -322.0,
|
| 1685 |
+
"loss": 0.0501,
|
| 1686 |
+
"rewards/accuracies": 1.0,
|
| 1687 |
+
"rewards/chosen": 2.046875,
|
| 1688 |
+
"rewards/margins": 6.4375,
|
| 1689 |
+
"rewards/rejected": -4.40625,
|
| 1690 |
+
"step": 1100
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 1.7788461538461537,
|
| 1694 |
+
"grad_norm": 32.08142550351567,
|
| 1695 |
+
"learning_rate": 2.2624703087885984e-07,
|
| 1696 |
+
"logits/chosen": -0.17578125,
|
| 1697 |
+
"logits/rejected": -0.373046875,
|
| 1698 |
+
"logps/chosen": -356.0,
|
| 1699 |
+
"logps/rejected": -348.0,
|
| 1700 |
+
"loss": 0.0622,
|
| 1701 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1702 |
+
"rewards/chosen": 2.0625,
|
| 1703 |
+
"rewards/margins": 6.3125,
|
| 1704 |
+
"rewards/rejected": -4.25,
|
| 1705 |
+
"step": 1110
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 1.7948717948717947,
|
| 1709 |
+
"grad_norm": 10.94873802284618,
|
| 1710 |
+
"learning_rate": 2.2327790973871732e-07,
|
| 1711 |
+
"logits/chosen": -0.296875,
|
| 1712 |
+
"logits/rejected": -0.474609375,
|
| 1713 |
+
"logps/chosen": -372.0,
|
| 1714 |
+
"logps/rejected": -368.0,
|
| 1715 |
+
"loss": 0.0631,
|
| 1716 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1717 |
+
"rewards/chosen": 2.015625,
|
| 1718 |
+
"rewards/margins": 6.15625,
|
| 1719 |
+
"rewards/rejected": -4.125,
|
| 1720 |
+
"step": 1120
|
| 1721 |
+
},
|
| 1722 |
+
{
|
| 1723 |
+
"epoch": 1.810897435897436,
|
| 1724 |
+
"grad_norm": 4.164714441310206,
|
| 1725 |
+
"learning_rate": 2.2030878859857483e-07,
|
| 1726 |
+
"logits/chosen": -0.251953125,
|
| 1727 |
+
"logits/rejected": -0.330078125,
|
| 1728 |
+
"logps/chosen": -362.0,
|
| 1729 |
+
"logps/rejected": -346.0,
|
| 1730 |
+
"loss": 0.0571,
|
| 1731 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1732 |
+
"rewards/chosen": 2.015625,
|
| 1733 |
+
"rewards/margins": 6.21875,
|
| 1734 |
+
"rewards/rejected": -4.1875,
|
| 1735 |
+
"step": 1130
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 1.8269230769230769,
|
| 1739 |
+
"grad_norm": 6.828795713228818,
|
| 1740 |
+
"learning_rate": 2.173396674584323e-07,
|
| 1741 |
+
"logits/chosen": -0.234375,
|
| 1742 |
+
"logits/rejected": -0.33984375,
|
| 1743 |
+
"logps/chosen": -376.0,
|
| 1744 |
+
"logps/rejected": -352.0,
|
| 1745 |
+
"loss": 0.1227,
|
| 1746 |
+
"rewards/accuracies": 0.949999988079071,
|
| 1747 |
+
"rewards/chosen": 2.234375,
|
| 1748 |
+
"rewards/margins": 5.5,
|
| 1749 |
+
"rewards/rejected": -3.234375,
|
| 1750 |
+
"step": 1140
|
| 1751 |
+
},
|
| 1752 |
+
{
|
| 1753 |
+
"epoch": 1.842948717948718,
|
| 1754 |
+
"grad_norm": 34.413780149566165,
|
| 1755 |
+
"learning_rate": 2.1437054631828976e-07,
|
| 1756 |
+
"logits/chosen": -0.224609375,
|
| 1757 |
+
"logits/rejected": -0.296875,
|
| 1758 |
+
"logps/chosen": -390.0,
|
| 1759 |
+
"logps/rejected": -346.0,
|
| 1760 |
+
"loss": 0.0492,
|
| 1761 |
+
"rewards/accuracies": 1.0,
|
| 1762 |
+
"rewards/chosen": 1.640625,
|
| 1763 |
+
"rewards/margins": 6.1875,
|
| 1764 |
+
"rewards/rejected": -4.5625,
|
| 1765 |
+
"step": 1150
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 1.858974358974359,
|
| 1769 |
+
"grad_norm": 3.332529440650818,
|
| 1770 |
+
"learning_rate": 2.1140142517814726e-07,
|
| 1771 |
+
"logits/chosen": -0.1669921875,
|
| 1772 |
+
"logits/rejected": -0.376953125,
|
| 1773 |
+
"logps/chosen": -352.0,
|
| 1774 |
+
"logps/rejected": -370.0,
|
| 1775 |
+
"loss": 0.0768,
|
| 1776 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1777 |
+
"rewards/chosen": 1.875,
|
| 1778 |
+
"rewards/margins": 6.34375,
|
| 1779 |
+
"rewards/rejected": -4.46875,
|
| 1780 |
+
"step": 1160
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 1.875,
|
| 1784 |
+
"grad_norm": 8.41589370404434,
|
| 1785 |
+
"learning_rate": 2.0843230403800474e-07,
|
| 1786 |
+
"logits/chosen": -0.349609375,
|
| 1787 |
+
"logits/rejected": -0.31640625,
|
| 1788 |
+
"logps/chosen": -374.0,
|
| 1789 |
+
"logps/rejected": -350.0,
|
| 1790 |
+
"loss": 0.0811,
|
| 1791 |
+
"rewards/accuracies": 0.9750000238418579,
|
| 1792 |
+
"rewards/chosen": 1.7578125,
|
| 1793 |
+
"rewards/margins": 5.5,
|
| 1794 |
+
"rewards/rejected": -3.75,
|
| 1795 |
+
"step": 1170
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 1.891025641025641,
|
| 1799 |
+
"grad_norm": 40.47797504678685,
|
| 1800 |
+
"learning_rate": 2.0546318289786222e-07,
|
| 1801 |
+
"logits/chosen": -0.2041015625,
|
| 1802 |
+
"logits/rejected": -0.40234375,
|
| 1803 |
+
"logps/chosen": -354.0,
|
| 1804 |
+
"logps/rejected": -366.0,
|
| 1805 |
+
"loss": 0.1088,
|
| 1806 |
+
"rewards/accuracies": 0.9375,
|
| 1807 |
+
"rewards/chosen": 1.5703125,
|
| 1808 |
+
"rewards/margins": 5.53125,
|
| 1809 |
+
"rewards/rejected": -3.953125,
|
| 1810 |
+
"step": 1180
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 1.907051282051282,
|
| 1814 |
+
"grad_norm": 8.055421977487784,
|
| 1815 |
+
"learning_rate": 2.0249406175771972e-07,
|
| 1816 |
+
"logits/chosen": -0.298828125,
|
| 1817 |
+
"logits/rejected": -0.482421875,
|
| 1818 |
+
"logps/chosen": -380.0,
|
| 1819 |
+
"logps/rejected": -354.0,
|
| 1820 |
+
"loss": 0.0963,
|
| 1821 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1822 |
+
"rewards/chosen": 2.046875,
|
| 1823 |
+
"rewards/margins": 6.0625,
|
| 1824 |
+
"rewards/rejected": -4.0,
|
| 1825 |
+
"step": 1190
|
| 1826 |
+
},
|
| 1827 |
+
{
|
| 1828 |
+
"epoch": 1.9230769230769231,
|
| 1829 |
+
"grad_norm": 1.0123224706730725,
|
| 1830 |
+
"learning_rate": 1.9952494061757718e-07,
|
| 1831 |
+
"logits/chosen": -0.259765625,
|
| 1832 |
+
"logits/rejected": -0.10888671875,
|
| 1833 |
+
"logps/chosen": -348.0,
|
| 1834 |
+
"logps/rejected": -356.0,
|
| 1835 |
+
"loss": 0.0319,
|
| 1836 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1837 |
+
"rewards/chosen": 2.71875,
|
| 1838 |
+
"rewards/margins": 7.03125,
|
| 1839 |
+
"rewards/rejected": -4.3125,
|
| 1840 |
+
"step": 1200
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 1.939102564102564,
|
| 1844 |
+
"grad_norm": 44.45376114208627,
|
| 1845 |
+
"learning_rate": 1.9655581947743466e-07,
|
| 1846 |
+
"logits/chosen": -0.21875,
|
| 1847 |
+
"logits/rejected": -0.3671875,
|
| 1848 |
+
"logps/chosen": -372.0,
|
| 1849 |
+
"logps/rejected": -376.0,
|
| 1850 |
+
"loss": 0.0746,
|
| 1851 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1852 |
+
"rewards/chosen": 1.7578125,
|
| 1853 |
+
"rewards/margins": 6.59375,
|
| 1854 |
+
"rewards/rejected": -4.8125,
|
| 1855 |
+
"step": 1210
|
| 1856 |
+
},
|
| 1857 |
+
{
|
| 1858 |
+
"epoch": 1.9551282051282053,
|
| 1859 |
+
"grad_norm": 15.603585643576269,
|
| 1860 |
+
"learning_rate": 1.9358669833729216e-07,
|
| 1861 |
+
"logits/chosen": -0.15234375,
|
| 1862 |
+
"logits/rejected": -0.373046875,
|
| 1863 |
+
"logps/chosen": -380.0,
|
| 1864 |
+
"logps/rejected": -352.0,
|
| 1865 |
+
"loss": 0.0608,
|
| 1866 |
+
"rewards/accuracies": 0.987500011920929,
|
| 1867 |
+
"rewards/chosen": 1.734375,
|
| 1868 |
+
"rewards/margins": 6.0625,
|
| 1869 |
+
"rewards/rejected": -4.3125,
|
| 1870 |
+
"step": 1220
|
| 1871 |
+
},
|
| 1872 |
+
{
|
| 1873 |
+
"epoch": 1.9711538461538463,
|
| 1874 |
+
"grad_norm": 13.268798051575008,
|
| 1875 |
+
"learning_rate": 1.9061757719714964e-07,
|
| 1876 |
+
"logits/chosen": -0.380859375,
|
| 1877 |
+
"logits/rejected": -0.45703125,
|
| 1878 |
+
"logps/chosen": -396.0,
|
| 1879 |
+
"logps/rejected": -362.0,
|
| 1880 |
+
"loss": 0.0417,
|
| 1881 |
+
"rewards/accuracies": 1.0,
|
| 1882 |
+
"rewards/chosen": 2.328125,
|
| 1883 |
+
"rewards/margins": 6.65625,
|
| 1884 |
+
"rewards/rejected": -4.3125,
|
| 1885 |
+
"step": 1230
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 1.9871794871794872,
|
| 1889 |
+
"grad_norm": 18.53274384798147,
|
| 1890 |
+
"learning_rate": 1.876484560570071e-07,
|
| 1891 |
+
"logits/chosen": -0.1357421875,
|
| 1892 |
+
"logits/rejected": -0.251953125,
|
| 1893 |
+
"logps/chosen": -390.0,
|
| 1894 |
+
"logps/rejected": -342.0,
|
| 1895 |
+
"loss": 0.0639,
|
| 1896 |
+
"rewards/accuracies": 0.9624999761581421,
|
| 1897 |
+
"rewards/chosen": 2.53125,
|
| 1898 |
+
"rewards/margins": 6.78125,
|
| 1899 |
+
"rewards/rejected": -4.25,
|
| 1900 |
+
"step": 1240
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 2.0,
|
| 1904 |
+
"eval_logits/chosen": -0.25,
|
| 1905 |
+
"eval_logits/rejected": -0.30078125,
|
| 1906 |
+
"eval_logps/chosen": -362.0,
|
| 1907 |
+
"eval_logps/rejected": -358.0,
|
| 1908 |
+
"eval_loss": 0.2055511474609375,
|
| 1909 |
+
"eval_rewards/accuracies": 0.9038461446762085,
|
| 1910 |
+
"eval_rewards/chosen": 1.859375,
|
| 1911 |
+
"eval_rewards/margins": 5.4375,
|
| 1912 |
+
"eval_rewards/rejected": -3.5625,
|
| 1913 |
+
"eval_runtime": 29.021,
|
| 1914 |
+
"eval_samples_per_second": 6.892,
|
| 1915 |
+
"eval_steps_per_second": 0.448,
|
| 1916 |
+
"step": 1248
|
| 1917 |
}
|
| 1918 |
],
|
| 1919 |
"logging_steps": 10,
|