Upload folder using huggingface_hub
Browse files- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr0.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr100.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr125.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr150.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr175.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr200.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr25.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr50.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr75.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_trainer_state.json +742 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr0.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr100.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr125.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr150.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr175.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr200.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr25.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr50.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr75.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_trainer_state.json +742 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr0.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr100.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr125.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr150.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr175.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr200.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr25.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr50.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr75.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_trainer_state.json +742 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr0.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr100.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr125.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr150.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr175.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr200.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr25.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr50.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr75.pth +3 -0
- client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_trainer_state.json +742 -0
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:41386a5932bd77be5f17f9dc9341984819b00771a125fc78fbaae4a3dda0c3a6
|
| 3 |
+
size 389170122
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9c624bf45c524486f29263548953f7aed0755d450fdb8a580f48859b622b86ba
|
| 3 |
+
size 389172166
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr100.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ae47bee9b6571e91ce6130dd12253d357c3ce1142bf01839c65e16fd07f869b
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr125.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:df252ebe33aa5cde72ea5e948dd71356613044af8212dcbdbbb9f8e90416f145
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr150.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7786d1e04a273bf982cebe4d7eee0bf45bad553632bce4ec1ed0a55cc323cdb7
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr175.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b224bda31bca245880fb998f06e16c7646ad73d2ba3c0508d87cd8350b65f25
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr200.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6d24c49c743b01a1546c802193e9d0da27b69ceda4c3127d03732a54d37d69f9
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr25.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:72a7a035d555a82a195e2a1a1bd82661cecf43a68b38fc2ac6b611c764684338
|
| 3 |
+
size 389172562
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr50.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cdc9e441da61a7f379ec858163739327cc287b0508cf683f109f673de2584841
|
| 3 |
+
size 389172562
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_client_model_round1_itr75.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c238f5cb59a017e26d18a16da9b3ba6632d66c2c82bb03158c2db34a5d149e95
|
| 3 |
+
size 389172562
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/0_trainer_state.json
ADDED
|
@@ -0,0 +1,742 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 201,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.009950248756218905,
|
| 13 |
+
"grad_norm": 1.7269937992095947,
|
| 14 |
+
"learning_rate": 1e-05,
|
| 15 |
+
"loss": 1.8672,
|
| 16 |
+
"step": 2
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01990049751243781,
|
| 20 |
+
"grad_norm": 1.523618221282959,
|
| 21 |
+
"learning_rate": 1e-05,
|
| 22 |
+
"loss": 1.7637,
|
| 23 |
+
"step": 4
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.029850746268656716,
|
| 27 |
+
"grad_norm": 0.9846981763839722,
|
| 28 |
+
"learning_rate": 1e-05,
|
| 29 |
+
"loss": 1.6396,
|
| 30 |
+
"step": 6
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.03980099502487562,
|
| 34 |
+
"grad_norm": 1.25033438205719,
|
| 35 |
+
"learning_rate": 1e-05,
|
| 36 |
+
"loss": 1.7197,
|
| 37 |
+
"step": 8
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04975124378109453,
|
| 41 |
+
"grad_norm": 1.5804775953292847,
|
| 42 |
+
"learning_rate": 1e-05,
|
| 43 |
+
"loss": 1.748,
|
| 44 |
+
"step": 10
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.05970149253731343,
|
| 48 |
+
"grad_norm": 1.946245551109314,
|
| 49 |
+
"learning_rate": 1e-05,
|
| 50 |
+
"loss": 2.2471,
|
| 51 |
+
"step": 12
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06965174129353234,
|
| 55 |
+
"grad_norm": 1.8690720796585083,
|
| 56 |
+
"learning_rate": 1e-05,
|
| 57 |
+
"loss": 1.6875,
|
| 58 |
+
"step": 14
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.07960199004975124,
|
| 62 |
+
"grad_norm": 2.4617319107055664,
|
| 63 |
+
"learning_rate": 1e-05,
|
| 64 |
+
"loss": 2.2178,
|
| 65 |
+
"step": 16
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.08955223880597014,
|
| 69 |
+
"grad_norm": 2.4414963722229004,
|
| 70 |
+
"learning_rate": 1e-05,
|
| 71 |
+
"loss": 2.2686,
|
| 72 |
+
"step": 18
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.09950248756218906,
|
| 76 |
+
"grad_norm": 1.1106815338134766,
|
| 77 |
+
"learning_rate": 1e-05,
|
| 78 |
+
"loss": 1.8477,
|
| 79 |
+
"step": 20
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.10945273631840796,
|
| 83 |
+
"grad_norm": 1.2677786350250244,
|
| 84 |
+
"learning_rate": 1e-05,
|
| 85 |
+
"loss": 1.8057,
|
| 86 |
+
"step": 22
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.11940298507462686,
|
| 90 |
+
"grad_norm": 0.9167836308479309,
|
| 91 |
+
"learning_rate": 1e-05,
|
| 92 |
+
"loss": 1.8877,
|
| 93 |
+
"step": 24
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.12935323383084577,
|
| 97 |
+
"grad_norm": 2.1515042781829834,
|
| 98 |
+
"learning_rate": 1e-05,
|
| 99 |
+
"loss": 1.6133,
|
| 100 |
+
"step": 26
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.13930348258706468,
|
| 104 |
+
"grad_norm": 0.6725850701332092,
|
| 105 |
+
"learning_rate": 1e-05,
|
| 106 |
+
"loss": 1.751,
|
| 107 |
+
"step": 28
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.14925373134328357,
|
| 111 |
+
"grad_norm": 1.40232253074646,
|
| 112 |
+
"learning_rate": 1e-05,
|
| 113 |
+
"loss": 1.6338,
|
| 114 |
+
"step": 30
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.15920398009950248,
|
| 118 |
+
"grad_norm": 1.2356222867965698,
|
| 119 |
+
"learning_rate": 1e-05,
|
| 120 |
+
"loss": 1.9746,
|
| 121 |
+
"step": 32
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1691542288557214,
|
| 125 |
+
"grad_norm": 1.3215711116790771,
|
| 126 |
+
"learning_rate": 1e-05,
|
| 127 |
+
"loss": 2.0088,
|
| 128 |
+
"step": 34
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1791044776119403,
|
| 132 |
+
"grad_norm": 1.6110485792160034,
|
| 133 |
+
"learning_rate": 1e-05,
|
| 134 |
+
"loss": 1.8936,
|
| 135 |
+
"step": 36
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1890547263681592,
|
| 139 |
+
"grad_norm": 0.7765350341796875,
|
| 140 |
+
"learning_rate": 1e-05,
|
| 141 |
+
"loss": 1.8486,
|
| 142 |
+
"step": 38
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.19900497512437812,
|
| 146 |
+
"grad_norm": 0.8672953248023987,
|
| 147 |
+
"learning_rate": 1e-05,
|
| 148 |
+
"loss": 1.6553,
|
| 149 |
+
"step": 40
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.208955223880597,
|
| 153 |
+
"grad_norm": 1.308184027671814,
|
| 154 |
+
"learning_rate": 1e-05,
|
| 155 |
+
"loss": 1.9473,
|
| 156 |
+
"step": 42
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.21890547263681592,
|
| 160 |
+
"grad_norm": 0.879946768283844,
|
| 161 |
+
"learning_rate": 1e-05,
|
| 162 |
+
"loss": 1.7314,
|
| 163 |
+
"step": 44
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.22885572139303484,
|
| 167 |
+
"grad_norm": 0.9189466834068298,
|
| 168 |
+
"learning_rate": 1e-05,
|
| 169 |
+
"loss": 1.9033,
|
| 170 |
+
"step": 46
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.23880597014925373,
|
| 174 |
+
"grad_norm": 0.9723676443099976,
|
| 175 |
+
"learning_rate": 1e-05,
|
| 176 |
+
"loss": 1.9805,
|
| 177 |
+
"step": 48
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.24875621890547264,
|
| 181 |
+
"grad_norm": 1.1213092803955078,
|
| 182 |
+
"learning_rate": 1e-05,
|
| 183 |
+
"loss": 1.8711,
|
| 184 |
+
"step": 50
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.25870646766169153,
|
| 188 |
+
"grad_norm": 0.8188961744308472,
|
| 189 |
+
"learning_rate": 1e-05,
|
| 190 |
+
"loss": 1.7061,
|
| 191 |
+
"step": 52
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.26865671641791045,
|
| 195 |
+
"grad_norm": 1.1325562000274658,
|
| 196 |
+
"learning_rate": 1e-05,
|
| 197 |
+
"loss": 2.0059,
|
| 198 |
+
"step": 54
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.27860696517412936,
|
| 202 |
+
"grad_norm": 0.9424170255661011,
|
| 203 |
+
"learning_rate": 1e-05,
|
| 204 |
+
"loss": 1.9326,
|
| 205 |
+
"step": 56
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.2885572139303483,
|
| 209 |
+
"grad_norm": 0.7555878758430481,
|
| 210 |
+
"learning_rate": 1e-05,
|
| 211 |
+
"loss": 1.8467,
|
| 212 |
+
"step": 58
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.29850746268656714,
|
| 216 |
+
"grad_norm": 1.2935634851455688,
|
| 217 |
+
"learning_rate": 1e-05,
|
| 218 |
+
"loss": 1.876,
|
| 219 |
+
"step": 60
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.30845771144278605,
|
| 223 |
+
"grad_norm": 0.8373547792434692,
|
| 224 |
+
"learning_rate": 1e-05,
|
| 225 |
+
"loss": 1.9648,
|
| 226 |
+
"step": 62
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.31840796019900497,
|
| 230 |
+
"grad_norm": 0.4899231195449829,
|
| 231 |
+
"learning_rate": 1e-05,
|
| 232 |
+
"loss": 1.9531,
|
| 233 |
+
"step": 64
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.3283582089552239,
|
| 237 |
+
"grad_norm": 0.7321932911872864,
|
| 238 |
+
"learning_rate": 1e-05,
|
| 239 |
+
"loss": 1.835,
|
| 240 |
+
"step": 66
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.3383084577114428,
|
| 244 |
+
"grad_norm": 1.1717967987060547,
|
| 245 |
+
"learning_rate": 1e-05,
|
| 246 |
+
"loss": 1.877,
|
| 247 |
+
"step": 68
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.3482587064676617,
|
| 251 |
+
"grad_norm": 0.8343319892883301,
|
| 252 |
+
"learning_rate": 1e-05,
|
| 253 |
+
"loss": 1.8662,
|
| 254 |
+
"step": 70
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.3582089552238806,
|
| 258 |
+
"grad_norm": 0.706469714641571,
|
| 259 |
+
"learning_rate": 1e-05,
|
| 260 |
+
"loss": 1.8438,
|
| 261 |
+
"step": 72
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.3681592039800995,
|
| 265 |
+
"grad_norm": 0.6939525604248047,
|
| 266 |
+
"learning_rate": 1e-05,
|
| 267 |
+
"loss": 1.8389,
|
| 268 |
+
"step": 74
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3781094527363184,
|
| 272 |
+
"grad_norm": 0.9119101166725159,
|
| 273 |
+
"learning_rate": 1e-05,
|
| 274 |
+
"loss": 1.7295,
|
| 275 |
+
"step": 76
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.3880597014925373,
|
| 279 |
+
"grad_norm": 0.9622534513473511,
|
| 280 |
+
"learning_rate": 1e-05,
|
| 281 |
+
"loss": 1.832,
|
| 282 |
+
"step": 78
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.39800995024875624,
|
| 286 |
+
"grad_norm": 1.3791886568069458,
|
| 287 |
+
"learning_rate": 1e-05,
|
| 288 |
+
"loss": 1.9189,
|
| 289 |
+
"step": 80
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.4079601990049751,
|
| 293 |
+
"grad_norm": 0.6077687740325928,
|
| 294 |
+
"learning_rate": 1e-05,
|
| 295 |
+
"loss": 1.7383,
|
| 296 |
+
"step": 82
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.417910447761194,
|
| 300 |
+
"grad_norm": 0.927112877368927,
|
| 301 |
+
"learning_rate": 1e-05,
|
| 302 |
+
"loss": 1.7188,
|
| 303 |
+
"step": 84
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.42786069651741293,
|
| 307 |
+
"grad_norm": 0.978523850440979,
|
| 308 |
+
"learning_rate": 1e-05,
|
| 309 |
+
"loss": 1.7871,
|
| 310 |
+
"step": 86
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.43781094527363185,
|
| 314 |
+
"grad_norm": 1.021173119544983,
|
| 315 |
+
"learning_rate": 1e-05,
|
| 316 |
+
"loss": 1.8877,
|
| 317 |
+
"step": 88
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.44776119402985076,
|
| 321 |
+
"grad_norm": 0.5392937660217285,
|
| 322 |
+
"learning_rate": 1e-05,
|
| 323 |
+
"loss": 1.8008,
|
| 324 |
+
"step": 90
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.4577114427860697,
|
| 328 |
+
"grad_norm": 0.8013123273849487,
|
| 329 |
+
"learning_rate": 1e-05,
|
| 330 |
+
"loss": 1.8281,
|
| 331 |
+
"step": 92
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.46766169154228854,
|
| 335 |
+
"grad_norm": 0.5712742209434509,
|
| 336 |
+
"learning_rate": 1e-05,
|
| 337 |
+
"loss": 2.0166,
|
| 338 |
+
"step": 94
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.47761194029850745,
|
| 342 |
+
"grad_norm": 0.8616108894348145,
|
| 343 |
+
"learning_rate": 1e-05,
|
| 344 |
+
"loss": 1.7783,
|
| 345 |
+
"step": 96
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.48756218905472637,
|
| 349 |
+
"grad_norm": 1.097840428352356,
|
| 350 |
+
"learning_rate": 1e-05,
|
| 351 |
+
"loss": 1.8594,
|
| 352 |
+
"step": 98
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.4975124378109453,
|
| 356 |
+
"grad_norm": 1.0751780271530151,
|
| 357 |
+
"learning_rate": 1e-05,
|
| 358 |
+
"loss": 1.9307,
|
| 359 |
+
"step": 100
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.5074626865671642,
|
| 363 |
+
"grad_norm": 0.5459370017051697,
|
| 364 |
+
"learning_rate": 1e-05,
|
| 365 |
+
"loss": 1.8057,
|
| 366 |
+
"step": 102
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5174129353233831,
|
| 370 |
+
"grad_norm": 0.6532576680183411,
|
| 371 |
+
"learning_rate": 1e-05,
|
| 372 |
+
"loss": 1.9131,
|
| 373 |
+
"step": 104
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.527363184079602,
|
| 377 |
+
"grad_norm": 0.7531883120536804,
|
| 378 |
+
"learning_rate": 1e-05,
|
| 379 |
+
"loss": 1.8516,
|
| 380 |
+
"step": 106
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.5373134328358209,
|
| 384 |
+
"grad_norm": 0.8269037008285522,
|
| 385 |
+
"learning_rate": 1e-05,
|
| 386 |
+
"loss": 1.8066,
|
| 387 |
+
"step": 108
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.5472636815920398,
|
| 391 |
+
"grad_norm": 0.6867979168891907,
|
| 392 |
+
"learning_rate": 1e-05,
|
| 393 |
+
"loss": 1.8008,
|
| 394 |
+
"step": 110
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.5572139303482587,
|
| 398 |
+
"grad_norm": 1.0090690851211548,
|
| 399 |
+
"learning_rate": 1e-05,
|
| 400 |
+
"loss": 1.9824,
|
| 401 |
+
"step": 112
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.5671641791044776,
|
| 405 |
+
"grad_norm": 0.9395070672035217,
|
| 406 |
+
"learning_rate": 1e-05,
|
| 407 |
+
"loss": 1.9043,
|
| 408 |
+
"step": 114
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.5771144278606966,
|
| 412 |
+
"grad_norm": 0.6765204071998596,
|
| 413 |
+
"learning_rate": 1e-05,
|
| 414 |
+
"loss": 1.8203,
|
| 415 |
+
"step": 116
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5870646766169154,
|
| 419 |
+
"grad_norm": 0.7717094421386719,
|
| 420 |
+
"learning_rate": 1e-05,
|
| 421 |
+
"loss": 1.8301,
|
| 422 |
+
"step": 118
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.5970149253731343,
|
| 426 |
+
"grad_norm": 0.83310467004776,
|
| 427 |
+
"learning_rate": 1e-05,
|
| 428 |
+
"loss": 1.8564,
|
| 429 |
+
"step": 120
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.6069651741293532,
|
| 433 |
+
"grad_norm": 0.7585615515708923,
|
| 434 |
+
"learning_rate": 1e-05,
|
| 435 |
+
"loss": 1.8721,
|
| 436 |
+
"step": 122
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.6169154228855721,
|
| 440 |
+
"grad_norm": 0.8300605416297913,
|
| 441 |
+
"learning_rate": 1e-05,
|
| 442 |
+
"loss": 1.8457,
|
| 443 |
+
"step": 124
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.6268656716417911,
|
| 447 |
+
"grad_norm": 0.6867866516113281,
|
| 448 |
+
"learning_rate": 1e-05,
|
| 449 |
+
"loss": 1.7969,
|
| 450 |
+
"step": 126
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.6368159203980099,
|
| 454 |
+
"grad_norm": 0.7056549787521362,
|
| 455 |
+
"learning_rate": 1e-05,
|
| 456 |
+
"loss": 1.8848,
|
| 457 |
+
"step": 128
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.6467661691542289,
|
| 461 |
+
"grad_norm": 0.7721547484397888,
|
| 462 |
+
"learning_rate": 1e-05,
|
| 463 |
+
"loss": 1.8057,
|
| 464 |
+
"step": 130
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.6567164179104478,
|
| 468 |
+
"grad_norm": 0.6794809103012085,
|
| 469 |
+
"learning_rate": 1e-05,
|
| 470 |
+
"loss": 1.7764,
|
| 471 |
+
"step": 132
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.6666666666666666,
|
| 475 |
+
"grad_norm": 1.111794114112854,
|
| 476 |
+
"learning_rate": 1e-05,
|
| 477 |
+
"loss": 1.6504,
|
| 478 |
+
"step": 134
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.6766169154228856,
|
| 482 |
+
"grad_norm": 1.2783085107803345,
|
| 483 |
+
"learning_rate": 1e-05,
|
| 484 |
+
"loss": 2.0264,
|
| 485 |
+
"step": 136
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6865671641791045,
|
| 489 |
+
"grad_norm": 0.9989943504333496,
|
| 490 |
+
"learning_rate": 1e-05,
|
| 491 |
+
"loss": 1.9395,
|
| 492 |
+
"step": 138
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.6965174129353234,
|
| 496 |
+
"grad_norm": 1.6281826496124268,
|
| 497 |
+
"learning_rate": 1e-05,
|
| 498 |
+
"loss": 1.998,
|
| 499 |
+
"step": 140
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.7064676616915423,
|
| 503 |
+
"grad_norm": 0.986221432685852,
|
| 504 |
+
"learning_rate": 1e-05,
|
| 505 |
+
"loss": 1.9229,
|
| 506 |
+
"step": 142
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.7164179104477612,
|
| 510 |
+
"grad_norm": 0.8353629112243652,
|
| 511 |
+
"learning_rate": 1e-05,
|
| 512 |
+
"loss": 1.9961,
|
| 513 |
+
"step": 144
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.7263681592039801,
|
| 517 |
+
"grad_norm": 0.7755623459815979,
|
| 518 |
+
"learning_rate": 1e-05,
|
| 519 |
+
"loss": 1.875,
|
| 520 |
+
"step": 146
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.736318407960199,
|
| 524 |
+
"grad_norm": 1.159177303314209,
|
| 525 |
+
"learning_rate": 1e-05,
|
| 526 |
+
"loss": 1.9072,
|
| 527 |
+
"step": 148
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.746268656716418,
|
| 531 |
+
"grad_norm": 0.9620384573936462,
|
| 532 |
+
"learning_rate": 1e-05,
|
| 533 |
+
"loss": 1.8311,
|
| 534 |
+
"step": 150
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.7562189054726368,
|
| 538 |
+
"grad_norm": 0.9240114092826843,
|
| 539 |
+
"learning_rate": 1e-05,
|
| 540 |
+
"loss": 1.9297,
|
| 541 |
+
"step": 152
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.7661691542288557,
|
| 545 |
+
"grad_norm": 0.71659255027771,
|
| 546 |
+
"learning_rate": 1e-05,
|
| 547 |
+
"loss": 1.8154,
|
| 548 |
+
"step": 154
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.7761194029850746,
|
| 552 |
+
"grad_norm": 1.1734024286270142,
|
| 553 |
+
"learning_rate": 1e-05,
|
| 554 |
+
"loss": 1.7393,
|
| 555 |
+
"step": 156
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.7860696517412935,
|
| 559 |
+
"grad_norm": 0.7090504765510559,
|
| 560 |
+
"learning_rate": 1e-05,
|
| 561 |
+
"loss": 1.8125,
|
| 562 |
+
"step": 158
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.7960199004975125,
|
| 566 |
+
"grad_norm": 0.8130780458450317,
|
| 567 |
+
"learning_rate": 1e-05,
|
| 568 |
+
"loss": 1.8281,
|
| 569 |
+
"step": 160
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.8059701492537313,
|
| 573 |
+
"grad_norm": 1.0261595249176025,
|
| 574 |
+
"learning_rate": 1e-05,
|
| 575 |
+
"loss": 1.8467,
|
| 576 |
+
"step": 162
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.8159203980099502,
|
| 580 |
+
"grad_norm": 0.9244614839553833,
|
| 581 |
+
"learning_rate": 1e-05,
|
| 582 |
+
"loss": 2.126,
|
| 583 |
+
"step": 164
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.8258706467661692,
|
| 587 |
+
"grad_norm": 0.7791427373886108,
|
| 588 |
+
"learning_rate": 1e-05,
|
| 589 |
+
"loss": 1.8203,
|
| 590 |
+
"step": 166
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.835820895522388,
|
| 594 |
+
"grad_norm": 1.075839638710022,
|
| 595 |
+
"learning_rate": 1e-05,
|
| 596 |
+
"loss": 1.9082,
|
| 597 |
+
"step": 168
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.845771144278607,
|
| 601 |
+
"grad_norm": 0.8389611840248108,
|
| 602 |
+
"learning_rate": 1e-05,
|
| 603 |
+
"loss": 1.7168,
|
| 604 |
+
"step": 170
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.8557213930348259,
|
| 608 |
+
"grad_norm": 0.7603033185005188,
|
| 609 |
+
"learning_rate": 1e-05,
|
| 610 |
+
"loss": 1.8096,
|
| 611 |
+
"step": 172
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.8656716417910447,
|
| 615 |
+
"grad_norm": 0.7383372783660889,
|
| 616 |
+
"learning_rate": 1e-05,
|
| 617 |
+
"loss": 1.8525,
|
| 618 |
+
"step": 174
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.8756218905472637,
|
| 622 |
+
"grad_norm": 0.7991982698440552,
|
| 623 |
+
"learning_rate": 1e-05,
|
| 624 |
+
"loss": 1.8037,
|
| 625 |
+
"step": 176
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.8855721393034826,
|
| 629 |
+
"grad_norm": 0.6685608625411987,
|
| 630 |
+
"learning_rate": 1e-05,
|
| 631 |
+
"loss": 1.8203,
|
| 632 |
+
"step": 178
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.8955223880597015,
|
| 636 |
+
"grad_norm": 0.7487455010414124,
|
| 637 |
+
"learning_rate": 1e-05,
|
| 638 |
+
"loss": 1.7627,
|
| 639 |
+
"step": 180
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.9054726368159204,
|
| 643 |
+
"grad_norm": 0.8541538119316101,
|
| 644 |
+
"learning_rate": 1e-05,
|
| 645 |
+
"loss": 1.7461,
|
| 646 |
+
"step": 182
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.9154228855721394,
|
| 650 |
+
"grad_norm": 0.6473627090454102,
|
| 651 |
+
"learning_rate": 1e-05,
|
| 652 |
+
"loss": 1.7061,
|
| 653 |
+
"step": 184
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.9253731343283582,
|
| 657 |
+
"grad_norm": 0.803227424621582,
|
| 658 |
+
"learning_rate": 1e-05,
|
| 659 |
+
"loss": 1.8047,
|
| 660 |
+
"step": 186
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.9353233830845771,
|
| 664 |
+
"grad_norm": 0.5693659782409668,
|
| 665 |
+
"learning_rate": 1e-05,
|
| 666 |
+
"loss": 1.7754,
|
| 667 |
+
"step": 188
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.945273631840796,
|
| 671 |
+
"grad_norm": 0.7182290554046631,
|
| 672 |
+
"learning_rate": 1e-05,
|
| 673 |
+
"loss": 1.8477,
|
| 674 |
+
"step": 190
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.9552238805970149,
|
| 678 |
+
"grad_norm": 1.529331088066101,
|
| 679 |
+
"learning_rate": 1e-05,
|
| 680 |
+
"loss": 2.0439,
|
| 681 |
+
"step": 192
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.9651741293532339,
|
| 685 |
+
"grad_norm": 1.1493253707885742,
|
| 686 |
+
"learning_rate": 1e-05,
|
| 687 |
+
"loss": 1.8564,
|
| 688 |
+
"step": 194
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.9751243781094527,
|
| 692 |
+
"grad_norm": 1.3912551403045654,
|
| 693 |
+
"learning_rate": 1e-05,
|
| 694 |
+
"loss": 1.793,
|
| 695 |
+
"step": 196
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.9850746268656716,
|
| 699 |
+
"grad_norm": 1.627122402191162,
|
| 700 |
+
"learning_rate": 1e-05,
|
| 701 |
+
"loss": 1.7422,
|
| 702 |
+
"step": 198
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.9950248756218906,
|
| 706 |
+
"grad_norm": 1.6311967372894287,
|
| 707 |
+
"learning_rate": 1e-05,
|
| 708 |
+
"loss": 1.8154,
|
| 709 |
+
"step": 200
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.0,
|
| 713 |
+
"step": 201,
|
| 714 |
+
"total_flos": 4.207228707510682e+16,
|
| 715 |
+
"train_loss": 1.853515625,
|
| 716 |
+
"train_runtime": 546.1337,
|
| 717 |
+
"train_samples_per_second": 1.472,
|
| 718 |
+
"train_steps_per_second": 0.368
|
| 719 |
+
}
|
| 720 |
+
],
|
| 721 |
+
"logging_steps": 2,
|
| 722 |
+
"max_steps": 201,
|
| 723 |
+
"num_input_tokens_seen": 0,
|
| 724 |
+
"num_train_epochs": 1,
|
| 725 |
+
"save_steps": 500,
|
| 726 |
+
"stateful_callbacks": {
|
| 727 |
+
"TrainerControl": {
|
| 728 |
+
"args": {
|
| 729 |
+
"should_epoch_stop": false,
|
| 730 |
+
"should_evaluate": false,
|
| 731 |
+
"should_log": false,
|
| 732 |
+
"should_save": false,
|
| 733 |
+
"should_training_stop": false
|
| 734 |
+
},
|
| 735 |
+
"attributes": {}
|
| 736 |
+
}
|
| 737 |
+
},
|
| 738 |
+
"total_flos": 4.207228707510682e+16,
|
| 739 |
+
"train_batch_size": 1,
|
| 740 |
+
"trial_name": null,
|
| 741 |
+
"trial_params": null
|
| 742 |
+
}
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4f502cba030e42b3cb18d1774ee3079d2844930dda911741d19786c6c56572dc
|
| 3 |
+
size 389170122
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8a5c9042e706191b98a8294d518dd9fcd8affc72a50a2ef8f1a8c9c1c2b2a97f
|
| 3 |
+
size 389172166
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr100.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:90300cd9098aeda4ac4616b197f4f5b9bc89f96289630f729e906c3629618a8f
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr125.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3d66ed5f20354ccc5e869f4ecae73d752e46f284fbf48c253b26ea82e0ce8977
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr150.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:54625241592d5f035d56da2c0e65010a5efcf6314cb4664dfda944e0ad77c061
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr175.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:51bec36e6b24d0f7c89d0f1c2364c3bfe30ec42a433183942f03dd71ed52e3ed
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr200.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5cad29f66dae4f76c279776af04154a47cec4fbee6571dacc4ae99a8ef598ca4
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr25.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7a337642b19202e309c5f3b4479f62afc3c3cb916923ce72bcb945d93b70fb81
|
| 3 |
+
size 389172562
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr50.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:69af4d39babdca4dc9ac9e7ed77fcc20a484a701e72e4865f15d5b8b5cd4d517
|
| 3 |
+
size 389172562
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_client_model_round1_itr75.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:914d531cb4998707ef7c2bbb74d23c01210e56da57b2c4c2af694bc78b8b5997
|
| 3 |
+
size 389172562
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/1_trainer_state.json
ADDED
|
@@ -0,0 +1,742 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 201,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.009950248756218905,
|
| 13 |
+
"grad_norm": 2.3353095054626465,
|
| 14 |
+
"learning_rate": 1e-05,
|
| 15 |
+
"loss": 1.9287,
|
| 16 |
+
"step": 2
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01990049751243781,
|
| 20 |
+
"grad_norm": 2.2337381839752197,
|
| 21 |
+
"learning_rate": 1e-05,
|
| 22 |
+
"loss": 1.9824,
|
| 23 |
+
"step": 4
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.029850746268656716,
|
| 27 |
+
"grad_norm": 2.243337631225586,
|
| 28 |
+
"learning_rate": 1e-05,
|
| 29 |
+
"loss": 1.7266,
|
| 30 |
+
"step": 6
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.03980099502487562,
|
| 34 |
+
"grad_norm": 2.199084758758545,
|
| 35 |
+
"learning_rate": 1e-05,
|
| 36 |
+
"loss": 1.7842,
|
| 37 |
+
"step": 8
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04975124378109453,
|
| 41 |
+
"grad_norm": 2.1202964782714844,
|
| 42 |
+
"learning_rate": 1e-05,
|
| 43 |
+
"loss": 1.8389,
|
| 44 |
+
"step": 10
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.05970149253731343,
|
| 48 |
+
"grad_norm": 1.483475685119629,
|
| 49 |
+
"learning_rate": 1e-05,
|
| 50 |
+
"loss": 1.8682,
|
| 51 |
+
"step": 12
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06965174129353234,
|
| 55 |
+
"grad_norm": 1.4063441753387451,
|
| 56 |
+
"learning_rate": 1e-05,
|
| 57 |
+
"loss": 2.1279,
|
| 58 |
+
"step": 14
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.07960199004975124,
|
| 62 |
+
"grad_norm": 2.0367581844329834,
|
| 63 |
+
"learning_rate": 1e-05,
|
| 64 |
+
"loss": 1.7773,
|
| 65 |
+
"step": 16
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.08955223880597014,
|
| 69 |
+
"grad_norm": 2.4807190895080566,
|
| 70 |
+
"learning_rate": 1e-05,
|
| 71 |
+
"loss": 2.2002,
|
| 72 |
+
"step": 18
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.09950248756218906,
|
| 76 |
+
"grad_norm": 2.546799898147583,
|
| 77 |
+
"learning_rate": 1e-05,
|
| 78 |
+
"loss": 1.6777,
|
| 79 |
+
"step": 20
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.10945273631840796,
|
| 83 |
+
"grad_norm": 1.718347430229187,
|
| 84 |
+
"learning_rate": 1e-05,
|
| 85 |
+
"loss": 2.0352,
|
| 86 |
+
"step": 22
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.11940298507462686,
|
| 90 |
+
"grad_norm": 2.2302751541137695,
|
| 91 |
+
"learning_rate": 1e-05,
|
| 92 |
+
"loss": 1.7676,
|
| 93 |
+
"step": 24
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.12935323383084577,
|
| 97 |
+
"grad_norm": 2.415226936340332,
|
| 98 |
+
"learning_rate": 1e-05,
|
| 99 |
+
"loss": 1.8154,
|
| 100 |
+
"step": 26
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.13930348258706468,
|
| 104 |
+
"grad_norm": 1.205698013305664,
|
| 105 |
+
"learning_rate": 1e-05,
|
| 106 |
+
"loss": 1.7148,
|
| 107 |
+
"step": 28
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.14925373134328357,
|
| 111 |
+
"grad_norm": 1.7399098873138428,
|
| 112 |
+
"learning_rate": 1e-05,
|
| 113 |
+
"loss": 1.916,
|
| 114 |
+
"step": 30
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.15920398009950248,
|
| 118 |
+
"grad_norm": 2.86334228515625,
|
| 119 |
+
"learning_rate": 1e-05,
|
| 120 |
+
"loss": 1.5806,
|
| 121 |
+
"step": 32
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1691542288557214,
|
| 125 |
+
"grad_norm": 2.26713228225708,
|
| 126 |
+
"learning_rate": 1e-05,
|
| 127 |
+
"loss": 1.8062,
|
| 128 |
+
"step": 34
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1791044776119403,
|
| 132 |
+
"grad_norm": 2.62178373336792,
|
| 133 |
+
"learning_rate": 1e-05,
|
| 134 |
+
"loss": 1.7021,
|
| 135 |
+
"step": 36
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1890547263681592,
|
| 139 |
+
"grad_norm": 2.2772014141082764,
|
| 140 |
+
"learning_rate": 1e-05,
|
| 141 |
+
"loss": 1.9297,
|
| 142 |
+
"step": 38
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.19900497512437812,
|
| 146 |
+
"grad_norm": 2.197669506072998,
|
| 147 |
+
"learning_rate": 1e-05,
|
| 148 |
+
"loss": 1.5771,
|
| 149 |
+
"step": 40
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.208955223880597,
|
| 153 |
+
"grad_norm": 2.128422260284424,
|
| 154 |
+
"learning_rate": 1e-05,
|
| 155 |
+
"loss": 1.6509,
|
| 156 |
+
"step": 42
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.21890547263681592,
|
| 160 |
+
"grad_norm": 2.7602224349975586,
|
| 161 |
+
"learning_rate": 1e-05,
|
| 162 |
+
"loss": 1.4043,
|
| 163 |
+
"step": 44
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.22885572139303484,
|
| 167 |
+
"grad_norm": 2.8489017486572266,
|
| 168 |
+
"learning_rate": 1e-05,
|
| 169 |
+
"loss": 1.2402,
|
| 170 |
+
"step": 46
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.23880597014925373,
|
| 174 |
+
"grad_norm": 2.247918128967285,
|
| 175 |
+
"learning_rate": 1e-05,
|
| 176 |
+
"loss": 1.2756,
|
| 177 |
+
"step": 48
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.24875621890547264,
|
| 181 |
+
"grad_norm": 3.031099796295166,
|
| 182 |
+
"learning_rate": 1e-05,
|
| 183 |
+
"loss": 1.2048,
|
| 184 |
+
"step": 50
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.25870646766169153,
|
| 188 |
+
"grad_norm": 3.75704288482666,
|
| 189 |
+
"learning_rate": 1e-05,
|
| 190 |
+
"loss": 0.8513,
|
| 191 |
+
"step": 52
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.26865671641791045,
|
| 195 |
+
"grad_norm": 4.332769870758057,
|
| 196 |
+
"learning_rate": 1e-05,
|
| 197 |
+
"loss": 1.3391,
|
| 198 |
+
"step": 54
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.27860696517412936,
|
| 202 |
+
"grad_norm": 2.794023036956787,
|
| 203 |
+
"learning_rate": 1e-05,
|
| 204 |
+
"loss": 0.5743,
|
| 205 |
+
"step": 56
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.2885572139303483,
|
| 209 |
+
"grad_norm": 3.639652729034424,
|
| 210 |
+
"learning_rate": 1e-05,
|
| 211 |
+
"loss": 0.5374,
|
| 212 |
+
"step": 58
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.29850746268656714,
|
| 216 |
+
"grad_norm": 7.670974254608154,
|
| 217 |
+
"learning_rate": 1e-05,
|
| 218 |
+
"loss": 1.0212,
|
| 219 |
+
"step": 60
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.30845771144278605,
|
| 223 |
+
"grad_norm": 7.943337440490723,
|
| 224 |
+
"learning_rate": 1e-05,
|
| 225 |
+
"loss": 1.2049,
|
| 226 |
+
"step": 62
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.31840796019900497,
|
| 230 |
+
"grad_norm": 7.651623249053955,
|
| 231 |
+
"learning_rate": 1e-05,
|
| 232 |
+
"loss": 1.3462,
|
| 233 |
+
"step": 64
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.3283582089552239,
|
| 237 |
+
"grad_norm": 11.468204498291016,
|
| 238 |
+
"learning_rate": 1e-05,
|
| 239 |
+
"loss": 1.1195,
|
| 240 |
+
"step": 66
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.3383084577114428,
|
| 244 |
+
"grad_norm": 1.7127485275268555,
|
| 245 |
+
"learning_rate": 1e-05,
|
| 246 |
+
"loss": 0.666,
|
| 247 |
+
"step": 68
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.3482587064676617,
|
| 251 |
+
"grad_norm": 8.698264122009277,
|
| 252 |
+
"learning_rate": 1e-05,
|
| 253 |
+
"loss": 0.4033,
|
| 254 |
+
"step": 70
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.3582089552238806,
|
| 258 |
+
"grad_norm": 10.121297836303711,
|
| 259 |
+
"learning_rate": 1e-05,
|
| 260 |
+
"loss": 1.7218,
|
| 261 |
+
"step": 72
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.3681592039800995,
|
| 265 |
+
"grad_norm": 6.066502571105957,
|
| 266 |
+
"learning_rate": 1e-05,
|
| 267 |
+
"loss": 1.5778,
|
| 268 |
+
"step": 74
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3781094527363184,
|
| 272 |
+
"grad_norm": 6.911867141723633,
|
| 273 |
+
"learning_rate": 1e-05,
|
| 274 |
+
"loss": 0.7626,
|
| 275 |
+
"step": 76
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.3880597014925373,
|
| 279 |
+
"grad_norm": 5.1254191398620605,
|
| 280 |
+
"learning_rate": 1e-05,
|
| 281 |
+
"loss": 0.6415,
|
| 282 |
+
"step": 78
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.39800995024875624,
|
| 286 |
+
"grad_norm": 5.967613697052002,
|
| 287 |
+
"learning_rate": 1e-05,
|
| 288 |
+
"loss": 1.9813,
|
| 289 |
+
"step": 80
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.4079601990049751,
|
| 293 |
+
"grad_norm": 3.8617615699768066,
|
| 294 |
+
"learning_rate": 1e-05,
|
| 295 |
+
"loss": 1.2355,
|
| 296 |
+
"step": 82
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.417910447761194,
|
| 300 |
+
"grad_norm": 2.3301353454589844,
|
| 301 |
+
"learning_rate": 1e-05,
|
| 302 |
+
"loss": 0.7871,
|
| 303 |
+
"step": 84
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.42786069651741293,
|
| 307 |
+
"grad_norm": 5.576817512512207,
|
| 308 |
+
"learning_rate": 1e-05,
|
| 309 |
+
"loss": 0.5404,
|
| 310 |
+
"step": 86
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.43781094527363185,
|
| 314 |
+
"grad_norm": 5.3230462074279785,
|
| 315 |
+
"learning_rate": 1e-05,
|
| 316 |
+
"loss": 0.7805,
|
| 317 |
+
"step": 88
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.44776119402985076,
|
| 321 |
+
"grad_norm": 6.264516353607178,
|
| 322 |
+
"learning_rate": 1e-05,
|
| 323 |
+
"loss": 1.4429,
|
| 324 |
+
"step": 90
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.4577114427860697,
|
| 328 |
+
"grad_norm": 3.8405826091766357,
|
| 329 |
+
"learning_rate": 1e-05,
|
| 330 |
+
"loss": 0.9402,
|
| 331 |
+
"step": 92
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.46766169154228854,
|
| 335 |
+
"grad_norm": 5.410178184509277,
|
| 336 |
+
"learning_rate": 1e-05,
|
| 337 |
+
"loss": 1.2462,
|
| 338 |
+
"step": 94
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.47761194029850745,
|
| 342 |
+
"grad_norm": 4.210456371307373,
|
| 343 |
+
"learning_rate": 1e-05,
|
| 344 |
+
"loss": 0.7014,
|
| 345 |
+
"step": 96
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.48756218905472637,
|
| 349 |
+
"grad_norm": 3.781709909439087,
|
| 350 |
+
"learning_rate": 1e-05,
|
| 351 |
+
"loss": 0.7141,
|
| 352 |
+
"step": 98
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.4975124378109453,
|
| 356 |
+
"grad_norm": 2.771514892578125,
|
| 357 |
+
"learning_rate": 1e-05,
|
| 358 |
+
"loss": 0.3659,
|
| 359 |
+
"step": 100
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.5074626865671642,
|
| 363 |
+
"grad_norm": 1.846204400062561,
|
| 364 |
+
"learning_rate": 1e-05,
|
| 365 |
+
"loss": 0.7911,
|
| 366 |
+
"step": 102
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5174129353233831,
|
| 370 |
+
"grad_norm": 3.2960546016693115,
|
| 371 |
+
"learning_rate": 1e-05,
|
| 372 |
+
"loss": 0.2299,
|
| 373 |
+
"step": 104
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.527363184079602,
|
| 377 |
+
"grad_norm": 8.897141456604004,
|
| 378 |
+
"learning_rate": 1e-05,
|
| 379 |
+
"loss": 0.7877,
|
| 380 |
+
"step": 106
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.5373134328358209,
|
| 384 |
+
"grad_norm": 8.636371612548828,
|
| 385 |
+
"learning_rate": 1e-05,
|
| 386 |
+
"loss": 1.1858,
|
| 387 |
+
"step": 108
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.5472636815920398,
|
| 391 |
+
"grad_norm": 5.115262031555176,
|
| 392 |
+
"learning_rate": 1e-05,
|
| 393 |
+
"loss": 1.1414,
|
| 394 |
+
"step": 110
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.5572139303482587,
|
| 398 |
+
"grad_norm": 2.9564952850341797,
|
| 399 |
+
"learning_rate": 1e-05,
|
| 400 |
+
"loss": 0.1131,
|
| 401 |
+
"step": 112
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.5671641791044776,
|
| 405 |
+
"grad_norm": 0.8790170550346375,
|
| 406 |
+
"learning_rate": 1e-05,
|
| 407 |
+
"loss": 0.1663,
|
| 408 |
+
"step": 114
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.5771144278606966,
|
| 412 |
+
"grad_norm": 3.546847105026245,
|
| 413 |
+
"learning_rate": 1e-05,
|
| 414 |
+
"loss": 0.4768,
|
| 415 |
+
"step": 116
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5870646766169154,
|
| 419 |
+
"grad_norm": 17.845115661621094,
|
| 420 |
+
"learning_rate": 1e-05,
|
| 421 |
+
"loss": 0.9104,
|
| 422 |
+
"step": 118
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.5970149253731343,
|
| 426 |
+
"grad_norm": 5.8693389892578125,
|
| 427 |
+
"learning_rate": 1e-05,
|
| 428 |
+
"loss": 1.1533,
|
| 429 |
+
"step": 120
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.6069651741293532,
|
| 433 |
+
"grad_norm": 8.540258407592773,
|
| 434 |
+
"learning_rate": 1e-05,
|
| 435 |
+
"loss": 0.7337,
|
| 436 |
+
"step": 122
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.6169154228855721,
|
| 440 |
+
"grad_norm": 7.653767108917236,
|
| 441 |
+
"learning_rate": 1e-05,
|
| 442 |
+
"loss": 1.3642,
|
| 443 |
+
"step": 124
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.6268656716417911,
|
| 447 |
+
"grad_norm": 3.6479005813598633,
|
| 448 |
+
"learning_rate": 1e-05,
|
| 449 |
+
"loss": 0.9125,
|
| 450 |
+
"step": 126
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.6368159203980099,
|
| 454 |
+
"grad_norm": 4.225715637207031,
|
| 455 |
+
"learning_rate": 1e-05,
|
| 456 |
+
"loss": 0.5348,
|
| 457 |
+
"step": 128
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.6467661691542289,
|
| 461 |
+
"grad_norm": 7.47725248336792,
|
| 462 |
+
"learning_rate": 1e-05,
|
| 463 |
+
"loss": 1.0396,
|
| 464 |
+
"step": 130
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.6567164179104478,
|
| 468 |
+
"grad_norm": 6.278067588806152,
|
| 469 |
+
"learning_rate": 1e-05,
|
| 470 |
+
"loss": 0.9005,
|
| 471 |
+
"step": 132
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.6666666666666666,
|
| 475 |
+
"grad_norm": 5.174953937530518,
|
| 476 |
+
"learning_rate": 1e-05,
|
| 477 |
+
"loss": 0.9641,
|
| 478 |
+
"step": 134
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.6766169154228856,
|
| 482 |
+
"grad_norm": 1.1321791410446167,
|
| 483 |
+
"learning_rate": 1e-05,
|
| 484 |
+
"loss": 0.1846,
|
| 485 |
+
"step": 136
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6865671641791045,
|
| 489 |
+
"grad_norm": 4.456625461578369,
|
| 490 |
+
"learning_rate": 1e-05,
|
| 491 |
+
"loss": 0.7826,
|
| 492 |
+
"step": 138
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.6965174129353234,
|
| 496 |
+
"grad_norm": 5.005406856536865,
|
| 497 |
+
"learning_rate": 1e-05,
|
| 498 |
+
"loss": 0.7396,
|
| 499 |
+
"step": 140
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.7064676616915423,
|
| 503 |
+
"grad_norm": 4.427835941314697,
|
| 504 |
+
"learning_rate": 1e-05,
|
| 505 |
+
"loss": 0.8176,
|
| 506 |
+
"step": 142
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.7164179104477612,
|
| 510 |
+
"grad_norm": 1.187164306640625,
|
| 511 |
+
"learning_rate": 1e-05,
|
| 512 |
+
"loss": 0.1401,
|
| 513 |
+
"step": 144
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.7263681592039801,
|
| 517 |
+
"grad_norm": 1.0914889574050903,
|
| 518 |
+
"learning_rate": 1e-05,
|
| 519 |
+
"loss": 0.2176,
|
| 520 |
+
"step": 146
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.736318407960199,
|
| 524 |
+
"grad_norm": 1.5438717603683472,
|
| 525 |
+
"learning_rate": 1e-05,
|
| 526 |
+
"loss": 0.2587,
|
| 527 |
+
"step": 148
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.746268656716418,
|
| 531 |
+
"grad_norm": 9.395261764526367,
|
| 532 |
+
"learning_rate": 1e-05,
|
| 533 |
+
"loss": 0.721,
|
| 534 |
+
"step": 150
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.7562189054726368,
|
| 538 |
+
"grad_norm": 4.722625732421875,
|
| 539 |
+
"learning_rate": 1e-05,
|
| 540 |
+
"loss": 0.4016,
|
| 541 |
+
"step": 152
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.7661691542288557,
|
| 545 |
+
"grad_norm": 2.3324036598205566,
|
| 546 |
+
"learning_rate": 1e-05,
|
| 547 |
+
"loss": 0.4424,
|
| 548 |
+
"step": 154
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.7761194029850746,
|
| 552 |
+
"grad_norm": 1.9911407232284546,
|
| 553 |
+
"learning_rate": 1e-05,
|
| 554 |
+
"loss": 0.1308,
|
| 555 |
+
"step": 156
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.7860696517412935,
|
| 559 |
+
"grad_norm": 1.4599279165267944,
|
| 560 |
+
"learning_rate": 1e-05,
|
| 561 |
+
"loss": 1.0972,
|
| 562 |
+
"step": 158
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.7960199004975125,
|
| 566 |
+
"grad_norm": 3.829977512359619,
|
| 567 |
+
"learning_rate": 1e-05,
|
| 568 |
+
"loss": 0.3672,
|
| 569 |
+
"step": 160
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.8059701492537313,
|
| 573 |
+
"grad_norm": 4.650771617889404,
|
| 574 |
+
"learning_rate": 1e-05,
|
| 575 |
+
"loss": 0.3544,
|
| 576 |
+
"step": 162
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.8159203980099502,
|
| 580 |
+
"grad_norm": 4.149600982666016,
|
| 581 |
+
"learning_rate": 1e-05,
|
| 582 |
+
"loss": 0.1115,
|
| 583 |
+
"step": 164
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.8258706467661692,
|
| 587 |
+
"grad_norm": 17.927310943603516,
|
| 588 |
+
"learning_rate": 1e-05,
|
| 589 |
+
"loss": 0.6831,
|
| 590 |
+
"step": 166
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.835820895522388,
|
| 594 |
+
"grad_norm": 5.11155366897583,
|
| 595 |
+
"learning_rate": 1e-05,
|
| 596 |
+
"loss": 0.2106,
|
| 597 |
+
"step": 168
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.845771144278607,
|
| 601 |
+
"grad_norm": 4.145267963409424,
|
| 602 |
+
"learning_rate": 1e-05,
|
| 603 |
+
"loss": 1.242,
|
| 604 |
+
"step": 170
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.8557213930348259,
|
| 608 |
+
"grad_norm": 2.1051363945007324,
|
| 609 |
+
"learning_rate": 1e-05,
|
| 610 |
+
"loss": 0.2696,
|
| 611 |
+
"step": 172
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.8656716417910447,
|
| 615 |
+
"grad_norm": 5.1266045570373535,
|
| 616 |
+
"learning_rate": 1e-05,
|
| 617 |
+
"loss": 0.7951,
|
| 618 |
+
"step": 174
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.8756218905472637,
|
| 622 |
+
"grad_norm": 0.3979372978210449,
|
| 623 |
+
"learning_rate": 1e-05,
|
| 624 |
+
"loss": 0.0854,
|
| 625 |
+
"step": 176
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.8855721393034826,
|
| 629 |
+
"grad_norm": 5.359825134277344,
|
| 630 |
+
"learning_rate": 1e-05,
|
| 631 |
+
"loss": 0.44,
|
| 632 |
+
"step": 178
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.8955223880597015,
|
| 636 |
+
"grad_norm": 0.5491272807121277,
|
| 637 |
+
"learning_rate": 1e-05,
|
| 638 |
+
"loss": 0.8416,
|
| 639 |
+
"step": 180
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.9054726368159204,
|
| 643 |
+
"grad_norm": 0.6715571284294128,
|
| 644 |
+
"learning_rate": 1e-05,
|
| 645 |
+
"loss": 0.0592,
|
| 646 |
+
"step": 182
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.9154228855721394,
|
| 650 |
+
"grad_norm": 1.3794856071472168,
|
| 651 |
+
"learning_rate": 1e-05,
|
| 652 |
+
"loss": 0.2191,
|
| 653 |
+
"step": 184
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.9253731343283582,
|
| 657 |
+
"grad_norm": 0.3467845916748047,
|
| 658 |
+
"learning_rate": 1e-05,
|
| 659 |
+
"loss": 0.0333,
|
| 660 |
+
"step": 186
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.9353233830845771,
|
| 664 |
+
"grad_norm": 3.0247490406036377,
|
| 665 |
+
"learning_rate": 1e-05,
|
| 666 |
+
"loss": 0.5178,
|
| 667 |
+
"step": 188
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.945273631840796,
|
| 671 |
+
"grad_norm": 2.2061026096343994,
|
| 672 |
+
"learning_rate": 1e-05,
|
| 673 |
+
"loss": 0.1849,
|
| 674 |
+
"step": 190
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.9552238805970149,
|
| 678 |
+
"grad_norm": 10.019746780395508,
|
| 679 |
+
"learning_rate": 1e-05,
|
| 680 |
+
"loss": 1.4295,
|
| 681 |
+
"step": 192
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.9651741293532339,
|
| 685 |
+
"grad_norm": 5.4788737297058105,
|
| 686 |
+
"learning_rate": 1e-05,
|
| 687 |
+
"loss": 0.4298,
|
| 688 |
+
"step": 194
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.9751243781094527,
|
| 692 |
+
"grad_norm": 2.0365283489227295,
|
| 693 |
+
"learning_rate": 1e-05,
|
| 694 |
+
"loss": 0.1699,
|
| 695 |
+
"step": 196
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.9850746268656716,
|
| 699 |
+
"grad_norm": 5.1401824951171875,
|
| 700 |
+
"learning_rate": 1e-05,
|
| 701 |
+
"loss": 0.6911,
|
| 702 |
+
"step": 198
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.9950248756218906,
|
| 706 |
+
"grad_norm": 9.342315673828125,
|
| 707 |
+
"learning_rate": 1e-05,
|
| 708 |
+
"loss": 1.1657,
|
| 709 |
+
"step": 200
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.0,
|
| 713 |
+
"step": 201,
|
| 714 |
+
"total_flos": 7.039446888428339e+16,
|
| 715 |
+
"train_loss": 0.9723156364402961,
|
| 716 |
+
"train_runtime": 918.2605,
|
| 717 |
+
"train_samples_per_second": 0.876,
|
| 718 |
+
"train_steps_per_second": 0.219
|
| 719 |
+
}
|
| 720 |
+
],
|
| 721 |
+
"logging_steps": 2,
|
| 722 |
+
"max_steps": 201,
|
| 723 |
+
"num_input_tokens_seen": 0,
|
| 724 |
+
"num_train_epochs": 1,
|
| 725 |
+
"save_steps": 500,
|
| 726 |
+
"stateful_callbacks": {
|
| 727 |
+
"TrainerControl": {
|
| 728 |
+
"args": {
|
| 729 |
+
"should_epoch_stop": false,
|
| 730 |
+
"should_evaluate": false,
|
| 731 |
+
"should_log": false,
|
| 732 |
+
"should_save": false,
|
| 733 |
+
"should_training_stop": false
|
| 734 |
+
},
|
| 735 |
+
"attributes": {}
|
| 736 |
+
}
|
| 737 |
+
},
|
| 738 |
+
"total_flos": 7.039446888428339e+16,
|
| 739 |
+
"train_batch_size": 1,
|
| 740 |
+
"trial_name": null,
|
| 741 |
+
"trial_params": null
|
| 742 |
+
}
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:afe0eb297d4b388ef6f3228dcad9fb8b4d9f685e5cf47d0ec142e39a8507e634
|
| 3 |
+
size 389170122
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:61ebb6c23a3e103919676ad06e8a5b44f804894a9e7932ba66716385a95db5cc
|
| 3 |
+
size 389172166
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr100.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:175e7b621c0a077878fc69929792934adfde16b37ccd49d8e72af1cacf91aa2a
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr125.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7900b1052ad8e857fc41ee627f9c48098c77c4a3ae0b859207a8fdfd468ea788
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr150.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f1ab7a85cabe9f59b527e8e18b0a4891e74d45d6ea078a37caa8951533bb4899
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr175.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f32a1d0d5ebfc02349b2949e1e2a52d55364dd7fc0c93a5960c5761df44a519c
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr200.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7c6a6a662abb854c4de54219e5126401e98cf58cd8ada0c999acd99bfe1daba5
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr25.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a8eca3745cb3964a8d4984f00d86c4eefcb6a9b8ba49fbfde1858aa2a0ef0e92
|
| 3 |
+
size 389172562
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr50.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5dee27803088a90511cedd6d4516f945db11f49060a0e58172effa9130140d3c
|
| 3 |
+
size 389172562
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_client_model_round1_itr75.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7b7efb328f1f760d9eb85ab7520e1c77f41d6c2ca5f31638e88227c694a3a882
|
| 3 |
+
size 389172562
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/2_trainer_state.json
ADDED
|
@@ -0,0 +1,742 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 201,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.009950248756218905,
|
| 13 |
+
"grad_norm": 2.9125924110412598,
|
| 14 |
+
"learning_rate": 1e-05,
|
| 15 |
+
"loss": 0.7976,
|
| 16 |
+
"step": 2
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01990049751243781,
|
| 20 |
+
"grad_norm": 10.521402359008789,
|
| 21 |
+
"learning_rate": 1e-05,
|
| 22 |
+
"loss": 1.7336,
|
| 23 |
+
"step": 4
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.029850746268656716,
|
| 27 |
+
"grad_norm": 2.8552589416503906,
|
| 28 |
+
"learning_rate": 1e-05,
|
| 29 |
+
"loss": 0.9961,
|
| 30 |
+
"step": 6
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.03980099502487562,
|
| 34 |
+
"grad_norm": 2.738194465637207,
|
| 35 |
+
"learning_rate": 1e-05,
|
| 36 |
+
"loss": 1.0359,
|
| 37 |
+
"step": 8
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04975124378109453,
|
| 41 |
+
"grad_norm": 2.78497576713562,
|
| 42 |
+
"learning_rate": 1e-05,
|
| 43 |
+
"loss": 0.9058,
|
| 44 |
+
"step": 10
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.05970149253731343,
|
| 48 |
+
"grad_norm": 2.7251193523406982,
|
| 49 |
+
"learning_rate": 1e-05,
|
| 50 |
+
"loss": 1.2095,
|
| 51 |
+
"step": 12
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06965174129353234,
|
| 55 |
+
"grad_norm": 2.718686103820801,
|
| 56 |
+
"learning_rate": 1e-05,
|
| 57 |
+
"loss": 1.3574,
|
| 58 |
+
"step": 14
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.07960199004975124,
|
| 62 |
+
"grad_norm": 2.7356369495391846,
|
| 63 |
+
"learning_rate": 1e-05,
|
| 64 |
+
"loss": 1.1289,
|
| 65 |
+
"step": 16
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.08955223880597014,
|
| 69 |
+
"grad_norm": 1.5454893112182617,
|
| 70 |
+
"learning_rate": 1e-05,
|
| 71 |
+
"loss": 0.7744,
|
| 72 |
+
"step": 18
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.09950248756218906,
|
| 76 |
+
"grad_norm": 1.5284944772720337,
|
| 77 |
+
"learning_rate": 1e-05,
|
| 78 |
+
"loss": 0.9192,
|
| 79 |
+
"step": 20
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.10945273631840796,
|
| 83 |
+
"grad_norm": 1.885312795639038,
|
| 84 |
+
"learning_rate": 1e-05,
|
| 85 |
+
"loss": 0.9624,
|
| 86 |
+
"step": 22
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.11940298507462686,
|
| 90 |
+
"grad_norm": 1.7953182458877563,
|
| 91 |
+
"learning_rate": 1e-05,
|
| 92 |
+
"loss": 1.0117,
|
| 93 |
+
"step": 24
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.12935323383084577,
|
| 97 |
+
"grad_norm": 1.7994076013565063,
|
| 98 |
+
"learning_rate": 1e-05,
|
| 99 |
+
"loss": 0.8989,
|
| 100 |
+
"step": 26
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.13930348258706468,
|
| 104 |
+
"grad_norm": 1.4220772981643677,
|
| 105 |
+
"learning_rate": 1e-05,
|
| 106 |
+
"loss": 0.9766,
|
| 107 |
+
"step": 28
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.14925373134328357,
|
| 111 |
+
"grad_norm": 1.1946817636489868,
|
| 112 |
+
"learning_rate": 1e-05,
|
| 113 |
+
"loss": 0.802,
|
| 114 |
+
"step": 30
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.15920398009950248,
|
| 118 |
+
"grad_norm": 2.1209020614624023,
|
| 119 |
+
"learning_rate": 1e-05,
|
| 120 |
+
"loss": 0.9346,
|
| 121 |
+
"step": 32
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1691542288557214,
|
| 125 |
+
"grad_norm": 1.3521199226379395,
|
| 126 |
+
"learning_rate": 1e-05,
|
| 127 |
+
"loss": 0.8032,
|
| 128 |
+
"step": 34
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1791044776119403,
|
| 132 |
+
"grad_norm": 2.0225729942321777,
|
| 133 |
+
"learning_rate": 1e-05,
|
| 134 |
+
"loss": 1.0701,
|
| 135 |
+
"step": 36
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1890547263681592,
|
| 139 |
+
"grad_norm": 0.9383809566497803,
|
| 140 |
+
"learning_rate": 1e-05,
|
| 141 |
+
"loss": 0.9634,
|
| 142 |
+
"step": 38
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.19900497512437812,
|
| 146 |
+
"grad_norm": 0.9488417506217957,
|
| 147 |
+
"learning_rate": 1e-05,
|
| 148 |
+
"loss": 0.802,
|
| 149 |
+
"step": 40
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.208955223880597,
|
| 153 |
+
"grad_norm": 1.0027282238006592,
|
| 154 |
+
"learning_rate": 1e-05,
|
| 155 |
+
"loss": 0.8672,
|
| 156 |
+
"step": 42
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.21890547263681592,
|
| 160 |
+
"grad_norm": 1.739571452140808,
|
| 161 |
+
"learning_rate": 1e-05,
|
| 162 |
+
"loss": 0.7917,
|
| 163 |
+
"step": 44
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.22885572139303484,
|
| 167 |
+
"grad_norm": 1.0754308700561523,
|
| 168 |
+
"learning_rate": 1e-05,
|
| 169 |
+
"loss": 0.9817,
|
| 170 |
+
"step": 46
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.23880597014925373,
|
| 174 |
+
"grad_norm": 2.2200374603271484,
|
| 175 |
+
"learning_rate": 1e-05,
|
| 176 |
+
"loss": 1.04,
|
| 177 |
+
"step": 48
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.24875621890547264,
|
| 181 |
+
"grad_norm": 2.2536051273345947,
|
| 182 |
+
"learning_rate": 1e-05,
|
| 183 |
+
"loss": 0.8911,
|
| 184 |
+
"step": 50
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.25870646766169153,
|
| 188 |
+
"grad_norm": 1.1104068756103516,
|
| 189 |
+
"learning_rate": 1e-05,
|
| 190 |
+
"loss": 0.7725,
|
| 191 |
+
"step": 52
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.26865671641791045,
|
| 195 |
+
"grad_norm": 0.8590786457061768,
|
| 196 |
+
"learning_rate": 1e-05,
|
| 197 |
+
"loss": 0.7488,
|
| 198 |
+
"step": 54
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.27860696517412936,
|
| 202 |
+
"grad_norm": 1.597180962562561,
|
| 203 |
+
"learning_rate": 1e-05,
|
| 204 |
+
"loss": 0.9021,
|
| 205 |
+
"step": 56
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.2885572139303483,
|
| 209 |
+
"grad_norm": 1.4583905935287476,
|
| 210 |
+
"learning_rate": 1e-05,
|
| 211 |
+
"loss": 1.1396,
|
| 212 |
+
"step": 58
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.29850746268656714,
|
| 216 |
+
"grad_norm": 2.604093313217163,
|
| 217 |
+
"learning_rate": 1e-05,
|
| 218 |
+
"loss": 1.4558,
|
| 219 |
+
"step": 60
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.30845771144278605,
|
| 223 |
+
"grad_norm": 1.1122664213180542,
|
| 224 |
+
"learning_rate": 1e-05,
|
| 225 |
+
"loss": 0.7021,
|
| 226 |
+
"step": 62
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.31840796019900497,
|
| 230 |
+
"grad_norm": 1.2909464836120605,
|
| 231 |
+
"learning_rate": 1e-05,
|
| 232 |
+
"loss": 1.1824,
|
| 233 |
+
"step": 64
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.3283582089552239,
|
| 237 |
+
"grad_norm": 1.6153855323791504,
|
| 238 |
+
"learning_rate": 1e-05,
|
| 239 |
+
"loss": 1.0962,
|
| 240 |
+
"step": 66
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.3383084577114428,
|
| 244 |
+
"grad_norm": 0.8087587356567383,
|
| 245 |
+
"learning_rate": 1e-05,
|
| 246 |
+
"loss": 1.0254,
|
| 247 |
+
"step": 68
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.3482587064676617,
|
| 251 |
+
"grad_norm": 0.988958477973938,
|
| 252 |
+
"learning_rate": 1e-05,
|
| 253 |
+
"loss": 0.9492,
|
| 254 |
+
"step": 70
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.3582089552238806,
|
| 258 |
+
"grad_norm": 0.9207158088684082,
|
| 259 |
+
"learning_rate": 1e-05,
|
| 260 |
+
"loss": 0.8652,
|
| 261 |
+
"step": 72
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.3681592039800995,
|
| 265 |
+
"grad_norm": 0.9971650242805481,
|
| 266 |
+
"learning_rate": 1e-05,
|
| 267 |
+
"loss": 0.9026,
|
| 268 |
+
"step": 74
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3781094527363184,
|
| 272 |
+
"grad_norm": 0.764649510383606,
|
| 273 |
+
"learning_rate": 1e-05,
|
| 274 |
+
"loss": 0.8691,
|
| 275 |
+
"step": 76
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.3880597014925373,
|
| 279 |
+
"grad_norm": 1.9914922714233398,
|
| 280 |
+
"learning_rate": 1e-05,
|
| 281 |
+
"loss": 1.2007,
|
| 282 |
+
"step": 78
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.39800995024875624,
|
| 286 |
+
"grad_norm": 1.3018932342529297,
|
| 287 |
+
"learning_rate": 1e-05,
|
| 288 |
+
"loss": 1.105,
|
| 289 |
+
"step": 80
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.4079601990049751,
|
| 293 |
+
"grad_norm": 0.9898261427879333,
|
| 294 |
+
"learning_rate": 1e-05,
|
| 295 |
+
"loss": 0.9951,
|
| 296 |
+
"step": 82
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.417910447761194,
|
| 300 |
+
"grad_norm": 2.1818137168884277,
|
| 301 |
+
"learning_rate": 1e-05,
|
| 302 |
+
"loss": 0.9082,
|
| 303 |
+
"step": 84
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.42786069651741293,
|
| 307 |
+
"grad_norm": 1.334093451499939,
|
| 308 |
+
"learning_rate": 1e-05,
|
| 309 |
+
"loss": 0.9263,
|
| 310 |
+
"step": 86
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.43781094527363185,
|
| 314 |
+
"grad_norm": 1.7600502967834473,
|
| 315 |
+
"learning_rate": 1e-05,
|
| 316 |
+
"loss": 1.085,
|
| 317 |
+
"step": 88
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.44776119402985076,
|
| 321 |
+
"grad_norm": 0.9962415099143982,
|
| 322 |
+
"learning_rate": 1e-05,
|
| 323 |
+
"loss": 0.8428,
|
| 324 |
+
"step": 90
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.4577114427860697,
|
| 328 |
+
"grad_norm": 0.6521207094192505,
|
| 329 |
+
"learning_rate": 1e-05,
|
| 330 |
+
"loss": 1.0254,
|
| 331 |
+
"step": 92
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.46766169154228854,
|
| 335 |
+
"grad_norm": 0.7943680882453918,
|
| 336 |
+
"learning_rate": 1e-05,
|
| 337 |
+
"loss": 0.877,
|
| 338 |
+
"step": 94
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.47761194029850745,
|
| 342 |
+
"grad_norm": 0.9639275074005127,
|
| 343 |
+
"learning_rate": 1e-05,
|
| 344 |
+
"loss": 0.9775,
|
| 345 |
+
"step": 96
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.48756218905472637,
|
| 349 |
+
"grad_norm": 0.5652957558631897,
|
| 350 |
+
"learning_rate": 1e-05,
|
| 351 |
+
"loss": 0.9736,
|
| 352 |
+
"step": 98
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.4975124378109453,
|
| 356 |
+
"grad_norm": 0.9656416773796082,
|
| 357 |
+
"learning_rate": 1e-05,
|
| 358 |
+
"loss": 0.9536,
|
| 359 |
+
"step": 100
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.5074626865671642,
|
| 363 |
+
"grad_norm": 1.2151691913604736,
|
| 364 |
+
"learning_rate": 1e-05,
|
| 365 |
+
"loss": 0.8682,
|
| 366 |
+
"step": 102
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5174129353233831,
|
| 370 |
+
"grad_norm": 0.8858196139335632,
|
| 371 |
+
"learning_rate": 1e-05,
|
| 372 |
+
"loss": 0.9375,
|
| 373 |
+
"step": 104
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.527363184079602,
|
| 377 |
+
"grad_norm": 0.6423347592353821,
|
| 378 |
+
"learning_rate": 1e-05,
|
| 379 |
+
"loss": 1.0142,
|
| 380 |
+
"step": 106
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.5373134328358209,
|
| 384 |
+
"grad_norm": 0.8421831727027893,
|
| 385 |
+
"learning_rate": 1e-05,
|
| 386 |
+
"loss": 0.9917,
|
| 387 |
+
"step": 108
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.5472636815920398,
|
| 391 |
+
"grad_norm": 0.4736503064632416,
|
| 392 |
+
"learning_rate": 1e-05,
|
| 393 |
+
"loss": 0.8921,
|
| 394 |
+
"step": 110
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.5572139303482587,
|
| 398 |
+
"grad_norm": 1.0089269876480103,
|
| 399 |
+
"learning_rate": 1e-05,
|
| 400 |
+
"loss": 0.9409,
|
| 401 |
+
"step": 112
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.5671641791044776,
|
| 405 |
+
"grad_norm": 0.7359282374382019,
|
| 406 |
+
"learning_rate": 1e-05,
|
| 407 |
+
"loss": 0.9326,
|
| 408 |
+
"step": 114
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.5771144278606966,
|
| 412 |
+
"grad_norm": 0.5314434170722961,
|
| 413 |
+
"learning_rate": 1e-05,
|
| 414 |
+
"loss": 0.958,
|
| 415 |
+
"step": 116
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5870646766169154,
|
| 419 |
+
"grad_norm": 1.1911386251449585,
|
| 420 |
+
"learning_rate": 1e-05,
|
| 421 |
+
"loss": 0.9512,
|
| 422 |
+
"step": 118
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.5970149253731343,
|
| 426 |
+
"grad_norm": 0.4896884858608246,
|
| 427 |
+
"learning_rate": 1e-05,
|
| 428 |
+
"loss": 0.8701,
|
| 429 |
+
"step": 120
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.6069651741293532,
|
| 433 |
+
"grad_norm": 0.5920294523239136,
|
| 434 |
+
"learning_rate": 1e-05,
|
| 435 |
+
"loss": 0.8062,
|
| 436 |
+
"step": 122
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.6169154228855721,
|
| 440 |
+
"grad_norm": 1.418971300125122,
|
| 441 |
+
"learning_rate": 1e-05,
|
| 442 |
+
"loss": 0.9956,
|
| 443 |
+
"step": 124
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.6268656716417911,
|
| 447 |
+
"grad_norm": 0.595043957233429,
|
| 448 |
+
"learning_rate": 1e-05,
|
| 449 |
+
"loss": 0.8882,
|
| 450 |
+
"step": 126
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.6368159203980099,
|
| 454 |
+
"grad_norm": 0.4718357026576996,
|
| 455 |
+
"learning_rate": 1e-05,
|
| 456 |
+
"loss": 1.0073,
|
| 457 |
+
"step": 128
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.6467661691542289,
|
| 461 |
+
"grad_norm": 0.6629860997200012,
|
| 462 |
+
"learning_rate": 1e-05,
|
| 463 |
+
"loss": 0.8784,
|
| 464 |
+
"step": 130
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.6567164179104478,
|
| 468 |
+
"grad_norm": 0.44892677664756775,
|
| 469 |
+
"learning_rate": 1e-05,
|
| 470 |
+
"loss": 0.9287,
|
| 471 |
+
"step": 132
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.6666666666666666,
|
| 475 |
+
"grad_norm": 1.1331515312194824,
|
| 476 |
+
"learning_rate": 1e-05,
|
| 477 |
+
"loss": 0.8667,
|
| 478 |
+
"step": 134
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.6766169154228856,
|
| 482 |
+
"grad_norm": 0.7462373375892639,
|
| 483 |
+
"learning_rate": 1e-05,
|
| 484 |
+
"loss": 0.8965,
|
| 485 |
+
"step": 136
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6865671641791045,
|
| 489 |
+
"grad_norm": 0.5195338726043701,
|
| 490 |
+
"learning_rate": 1e-05,
|
| 491 |
+
"loss": 0.937,
|
| 492 |
+
"step": 138
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.6965174129353234,
|
| 496 |
+
"grad_norm": 0.817085325717926,
|
| 497 |
+
"learning_rate": 1e-05,
|
| 498 |
+
"loss": 1.0107,
|
| 499 |
+
"step": 140
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.7064676616915423,
|
| 503 |
+
"grad_norm": 0.4965575039386749,
|
| 504 |
+
"learning_rate": 1e-05,
|
| 505 |
+
"loss": 0.9795,
|
| 506 |
+
"step": 142
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.7164179104477612,
|
| 510 |
+
"grad_norm": 0.47578611969947815,
|
| 511 |
+
"learning_rate": 1e-05,
|
| 512 |
+
"loss": 0.8784,
|
| 513 |
+
"step": 144
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.7263681592039801,
|
| 517 |
+
"grad_norm": 0.6801284551620483,
|
| 518 |
+
"learning_rate": 1e-05,
|
| 519 |
+
"loss": 0.8589,
|
| 520 |
+
"step": 146
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.736318407960199,
|
| 524 |
+
"grad_norm": 0.4817887544631958,
|
| 525 |
+
"learning_rate": 1e-05,
|
| 526 |
+
"loss": 0.8599,
|
| 527 |
+
"step": 148
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.746268656716418,
|
| 531 |
+
"grad_norm": 0.944343626499176,
|
| 532 |
+
"learning_rate": 1e-05,
|
| 533 |
+
"loss": 0.9556,
|
| 534 |
+
"step": 150
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.7562189054726368,
|
| 538 |
+
"grad_norm": 0.6698830127716064,
|
| 539 |
+
"learning_rate": 1e-05,
|
| 540 |
+
"loss": 0.8813,
|
| 541 |
+
"step": 152
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.7661691542288557,
|
| 545 |
+
"grad_norm": 0.4846048653125763,
|
| 546 |
+
"learning_rate": 1e-05,
|
| 547 |
+
"loss": 1.0513,
|
| 548 |
+
"step": 154
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.7761194029850746,
|
| 552 |
+
"grad_norm": 0.6287723183631897,
|
| 553 |
+
"learning_rate": 1e-05,
|
| 554 |
+
"loss": 0.9214,
|
| 555 |
+
"step": 156
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.7860696517412935,
|
| 559 |
+
"grad_norm": 0.49072515964508057,
|
| 560 |
+
"learning_rate": 1e-05,
|
| 561 |
+
"loss": 1.0171,
|
| 562 |
+
"step": 158
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.7960199004975125,
|
| 566 |
+
"grad_norm": 0.5004100799560547,
|
| 567 |
+
"learning_rate": 1e-05,
|
| 568 |
+
"loss": 0.9619,
|
| 569 |
+
"step": 160
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.8059701492537313,
|
| 573 |
+
"grad_norm": 0.8207651972770691,
|
| 574 |
+
"learning_rate": 1e-05,
|
| 575 |
+
"loss": 0.9321,
|
| 576 |
+
"step": 162
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.8159203980099502,
|
| 580 |
+
"grad_norm": 0.560549795627594,
|
| 581 |
+
"learning_rate": 1e-05,
|
| 582 |
+
"loss": 0.9092,
|
| 583 |
+
"step": 164
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.8258706467661692,
|
| 587 |
+
"grad_norm": 0.7873324155807495,
|
| 588 |
+
"learning_rate": 1e-05,
|
| 589 |
+
"loss": 0.9258,
|
| 590 |
+
"step": 166
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.835820895522388,
|
| 594 |
+
"grad_norm": 0.6655442714691162,
|
| 595 |
+
"learning_rate": 1e-05,
|
| 596 |
+
"loss": 0.8848,
|
| 597 |
+
"step": 168
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.845771144278607,
|
| 601 |
+
"grad_norm": 0.8962746858596802,
|
| 602 |
+
"learning_rate": 1e-05,
|
| 603 |
+
"loss": 0.9116,
|
| 604 |
+
"step": 170
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.8557213930348259,
|
| 608 |
+
"grad_norm": 0.9736960530281067,
|
| 609 |
+
"learning_rate": 1e-05,
|
| 610 |
+
"loss": 0.9761,
|
| 611 |
+
"step": 172
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.8656716417910447,
|
| 615 |
+
"grad_norm": 0.7160093188285828,
|
| 616 |
+
"learning_rate": 1e-05,
|
| 617 |
+
"loss": 0.8901,
|
| 618 |
+
"step": 174
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.8756218905472637,
|
| 622 |
+
"grad_norm": 1.4417303800582886,
|
| 623 |
+
"learning_rate": 1e-05,
|
| 624 |
+
"loss": 0.9692,
|
| 625 |
+
"step": 176
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.8855721393034826,
|
| 629 |
+
"grad_norm": 0.5898114442825317,
|
| 630 |
+
"learning_rate": 1e-05,
|
| 631 |
+
"loss": 0.9106,
|
| 632 |
+
"step": 178
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.8955223880597015,
|
| 636 |
+
"grad_norm": 0.6590920686721802,
|
| 637 |
+
"learning_rate": 1e-05,
|
| 638 |
+
"loss": 0.8313,
|
| 639 |
+
"step": 180
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.9054726368159204,
|
| 643 |
+
"grad_norm": 1.4728569984436035,
|
| 644 |
+
"learning_rate": 1e-05,
|
| 645 |
+
"loss": 0.9297,
|
| 646 |
+
"step": 182
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.9154228855721394,
|
| 650 |
+
"grad_norm": 0.7601069808006287,
|
| 651 |
+
"learning_rate": 1e-05,
|
| 652 |
+
"loss": 0.918,
|
| 653 |
+
"step": 184
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.9253731343283582,
|
| 657 |
+
"grad_norm": 0.637050449848175,
|
| 658 |
+
"learning_rate": 1e-05,
|
| 659 |
+
"loss": 0.8643,
|
| 660 |
+
"step": 186
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.9353233830845771,
|
| 664 |
+
"grad_norm": 0.9461324214935303,
|
| 665 |
+
"learning_rate": 1e-05,
|
| 666 |
+
"loss": 0.9092,
|
| 667 |
+
"step": 188
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.945273631840796,
|
| 671 |
+
"grad_norm": 0.6568467020988464,
|
| 672 |
+
"learning_rate": 1e-05,
|
| 673 |
+
"loss": 0.897,
|
| 674 |
+
"step": 190
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.9552238805970149,
|
| 678 |
+
"grad_norm": 1.0706391334533691,
|
| 679 |
+
"learning_rate": 1e-05,
|
| 680 |
+
"loss": 0.9268,
|
| 681 |
+
"step": 192
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.9651741293532339,
|
| 685 |
+
"grad_norm": 0.9642358422279358,
|
| 686 |
+
"learning_rate": 1e-05,
|
| 687 |
+
"loss": 0.9355,
|
| 688 |
+
"step": 194
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.9751243781094527,
|
| 692 |
+
"grad_norm": 1.1096305847167969,
|
| 693 |
+
"learning_rate": 1e-05,
|
| 694 |
+
"loss": 0.7712,
|
| 695 |
+
"step": 196
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.9850746268656716,
|
| 699 |
+
"grad_norm": 0.8058964610099792,
|
| 700 |
+
"learning_rate": 1e-05,
|
| 701 |
+
"loss": 0.9238,
|
| 702 |
+
"step": 198
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.9950248756218906,
|
| 706 |
+
"grad_norm": 1.3148071765899658,
|
| 707 |
+
"learning_rate": 1e-05,
|
| 708 |
+
"loss": 1.0073,
|
| 709 |
+
"step": 200
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.0,
|
| 713 |
+
"step": 201,
|
| 714 |
+
"total_flos": 1.9968928570671104e+16,
|
| 715 |
+
"train_loss": 0.9518885066853234,
|
| 716 |
+
"train_runtime": 446.5447,
|
| 717 |
+
"train_samples_per_second": 1.8,
|
| 718 |
+
"train_steps_per_second": 0.45
|
| 719 |
+
}
|
| 720 |
+
],
|
| 721 |
+
"logging_steps": 2,
|
| 722 |
+
"max_steps": 201,
|
| 723 |
+
"num_input_tokens_seen": 0,
|
| 724 |
+
"num_train_epochs": 1,
|
| 725 |
+
"save_steps": 500,
|
| 726 |
+
"stateful_callbacks": {
|
| 727 |
+
"TrainerControl": {
|
| 728 |
+
"args": {
|
| 729 |
+
"should_epoch_stop": false,
|
| 730 |
+
"should_evaluate": false,
|
| 731 |
+
"should_log": false,
|
| 732 |
+
"should_save": false,
|
| 733 |
+
"should_training_stop": false
|
| 734 |
+
},
|
| 735 |
+
"attributes": {}
|
| 736 |
+
}
|
| 737 |
+
},
|
| 738 |
+
"total_flos": 1.9968928570671104e+16,
|
| 739 |
+
"train_batch_size": 1,
|
| 740 |
+
"trial_name": null,
|
| 741 |
+
"trial_params": null
|
| 742 |
+
}
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33880b5b62e10da427e3da58e642d0ddf3180d5e6ee37b538f51a15c0615e985
|
| 3 |
+
size 389170122
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0381baaff4450f626197743fb5c383c3c3dba7b8082f3b8ebefd40d3a022152e
|
| 3 |
+
size 389172166
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr100.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b43c5b18ab0d0012dd785b53a4afb05dddb364cc6e9b9f3108f1f0be476fcc46
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr125.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8dcfd94c53330102d025a476ac79ed53451082069afa543295ec43d619bee258
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr150.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:75c2b5f5214c64af61f7e01c631c2828076f744eeba51e35ad7c7ecff7f82542
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr175.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3872a833de36d3cf4a3ea10a17ce3048c6a5e72256e7a48cfc83e8b4de12c8f
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr200.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7e629c074b7bf33dd5819a267bcbaf48d036fe06f51a3eac5f671f6af5fd5a64
|
| 3 |
+
size 389172958
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr25.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6e1285420c83a99d2ef09967a8472a22758b5874a981653c25f3434cb1c7f6c6
|
| 3 |
+
size 389172562
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr50.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:95454430cbc2efebf2e74e6984d06f7ed7d8a64ec04f109a8bf0653598f8bb80
|
| 3 |
+
size 389172562
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_client_model_round1_itr75.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:11234e37c6bf9150af86e746c3ad3e279f568fb971614d120f9ed0fb60c7c9fe
|
| 3 |
+
size 389172562
|
client_states_sft_bs4_saveoptim_lr1e-5_sc8_1tasks_1rounds_fixitr201_T0125_decay099_feddpa_r20/3_trainer_state.json
ADDED
|
@@ -0,0 +1,742 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 201,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.009950248756218905,
|
| 13 |
+
"grad_norm": 4.539283752441406,
|
| 14 |
+
"learning_rate": 1e-05,
|
| 15 |
+
"loss": 2.5386,
|
| 16 |
+
"step": 2
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01990049751243781,
|
| 20 |
+
"grad_norm": 6.657641410827637,
|
| 21 |
+
"learning_rate": 1e-05,
|
| 22 |
+
"loss": 2.8231,
|
| 23 |
+
"step": 4
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.029850746268656716,
|
| 27 |
+
"grad_norm": 3.6988651752471924,
|
| 28 |
+
"learning_rate": 1e-05,
|
| 29 |
+
"loss": 1.6801,
|
| 30 |
+
"step": 6
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.03980099502487562,
|
| 34 |
+
"grad_norm": 3.8296077251434326,
|
| 35 |
+
"learning_rate": 1e-05,
|
| 36 |
+
"loss": 1.1219,
|
| 37 |
+
"step": 8
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.04975124378109453,
|
| 41 |
+
"grad_norm": 8.80754280090332,
|
| 42 |
+
"learning_rate": 1e-05,
|
| 43 |
+
"loss": 2.1103,
|
| 44 |
+
"step": 10
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.05970149253731343,
|
| 48 |
+
"grad_norm": 6.436983585357666,
|
| 49 |
+
"learning_rate": 1e-05,
|
| 50 |
+
"loss": 2.0908,
|
| 51 |
+
"step": 12
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.06965174129353234,
|
| 55 |
+
"grad_norm": 5.299738883972168,
|
| 56 |
+
"learning_rate": 1e-05,
|
| 57 |
+
"loss": 1.6301,
|
| 58 |
+
"step": 14
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.07960199004975124,
|
| 62 |
+
"grad_norm": 4.144854545593262,
|
| 63 |
+
"learning_rate": 1e-05,
|
| 64 |
+
"loss": 1.4358,
|
| 65 |
+
"step": 16
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.08955223880597014,
|
| 69 |
+
"grad_norm": 5.3244709968566895,
|
| 70 |
+
"learning_rate": 1e-05,
|
| 71 |
+
"loss": 1.4669,
|
| 72 |
+
"step": 18
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.09950248756218906,
|
| 76 |
+
"grad_norm": 4.895822525024414,
|
| 77 |
+
"learning_rate": 1e-05,
|
| 78 |
+
"loss": 1.3648,
|
| 79 |
+
"step": 20
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.10945273631840796,
|
| 83 |
+
"grad_norm": 3.2509872913360596,
|
| 84 |
+
"learning_rate": 1e-05,
|
| 85 |
+
"loss": 0.9271,
|
| 86 |
+
"step": 22
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.11940298507462686,
|
| 90 |
+
"grad_norm": 3.2340168952941895,
|
| 91 |
+
"learning_rate": 1e-05,
|
| 92 |
+
"loss": 1.4258,
|
| 93 |
+
"step": 24
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.12935323383084577,
|
| 97 |
+
"grad_norm": 5.070857524871826,
|
| 98 |
+
"learning_rate": 1e-05,
|
| 99 |
+
"loss": 1.7946,
|
| 100 |
+
"step": 26
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.13930348258706468,
|
| 104 |
+
"grad_norm": 3.653397798538208,
|
| 105 |
+
"learning_rate": 1e-05,
|
| 106 |
+
"loss": 1.131,
|
| 107 |
+
"step": 28
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.14925373134328357,
|
| 111 |
+
"grad_norm": 5.358354091644287,
|
| 112 |
+
"learning_rate": 1e-05,
|
| 113 |
+
"loss": 1.6315,
|
| 114 |
+
"step": 30
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.15920398009950248,
|
| 118 |
+
"grad_norm": 1.0174429416656494,
|
| 119 |
+
"learning_rate": 1e-05,
|
| 120 |
+
"loss": 0.8319,
|
| 121 |
+
"step": 32
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1691542288557214,
|
| 125 |
+
"grad_norm": 4.7866668701171875,
|
| 126 |
+
"learning_rate": 1e-05,
|
| 127 |
+
"loss": 1.2824,
|
| 128 |
+
"step": 34
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.1791044776119403,
|
| 132 |
+
"grad_norm": 3.9776785373687744,
|
| 133 |
+
"learning_rate": 1e-05,
|
| 134 |
+
"loss": 1.4014,
|
| 135 |
+
"step": 36
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1890547263681592,
|
| 139 |
+
"grad_norm": 3.32464337348938,
|
| 140 |
+
"learning_rate": 1e-05,
|
| 141 |
+
"loss": 0.9191,
|
| 142 |
+
"step": 38
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.19900497512437812,
|
| 146 |
+
"grad_norm": 5.934720516204834,
|
| 147 |
+
"learning_rate": 1e-05,
|
| 148 |
+
"loss": 1.5676,
|
| 149 |
+
"step": 40
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.208955223880597,
|
| 153 |
+
"grad_norm": 4.061328887939453,
|
| 154 |
+
"learning_rate": 1e-05,
|
| 155 |
+
"loss": 0.9695,
|
| 156 |
+
"step": 42
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.21890547263681592,
|
| 160 |
+
"grad_norm": 3.4364242553710938,
|
| 161 |
+
"learning_rate": 1e-05,
|
| 162 |
+
"loss": 1.0359,
|
| 163 |
+
"step": 44
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.22885572139303484,
|
| 167 |
+
"grad_norm": 3.964568853378296,
|
| 168 |
+
"learning_rate": 1e-05,
|
| 169 |
+
"loss": 2.0101,
|
| 170 |
+
"step": 46
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.23880597014925373,
|
| 174 |
+
"grad_norm": 9.2665376663208,
|
| 175 |
+
"learning_rate": 1e-05,
|
| 176 |
+
"loss": 0.8134,
|
| 177 |
+
"step": 48
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.24875621890547264,
|
| 181 |
+
"grad_norm": 4.130026817321777,
|
| 182 |
+
"learning_rate": 1e-05,
|
| 183 |
+
"loss": 1.5751,
|
| 184 |
+
"step": 50
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.25870646766169153,
|
| 188 |
+
"grad_norm": 8.430376052856445,
|
| 189 |
+
"learning_rate": 1e-05,
|
| 190 |
+
"loss": 1.644,
|
| 191 |
+
"step": 52
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.26865671641791045,
|
| 195 |
+
"grad_norm": 3.5804402828216553,
|
| 196 |
+
"learning_rate": 1e-05,
|
| 197 |
+
"loss": 0.8242,
|
| 198 |
+
"step": 54
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.27860696517412936,
|
| 202 |
+
"grad_norm": 3.2127153873443604,
|
| 203 |
+
"learning_rate": 1e-05,
|
| 204 |
+
"loss": 1.1785,
|
| 205 |
+
"step": 56
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.2885572139303483,
|
| 209 |
+
"grad_norm": 7.125833034515381,
|
| 210 |
+
"learning_rate": 1e-05,
|
| 211 |
+
"loss": 1.3655,
|
| 212 |
+
"step": 58
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.29850746268656714,
|
| 216 |
+
"grad_norm": 8.425891876220703,
|
| 217 |
+
"learning_rate": 1e-05,
|
| 218 |
+
"loss": 2.6185,
|
| 219 |
+
"step": 60
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.30845771144278605,
|
| 223 |
+
"grad_norm": 2.727701187133789,
|
| 224 |
+
"learning_rate": 1e-05,
|
| 225 |
+
"loss": 0.602,
|
| 226 |
+
"step": 62
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.31840796019900497,
|
| 230 |
+
"grad_norm": 5.714324951171875,
|
| 231 |
+
"learning_rate": 1e-05,
|
| 232 |
+
"loss": 0.8775,
|
| 233 |
+
"step": 64
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.3283582089552239,
|
| 237 |
+
"grad_norm": 10.05617618560791,
|
| 238 |
+
"learning_rate": 1e-05,
|
| 239 |
+
"loss": 1.9613,
|
| 240 |
+
"step": 66
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.3383084577114428,
|
| 244 |
+
"grad_norm": 2.26175594329834,
|
| 245 |
+
"learning_rate": 1e-05,
|
| 246 |
+
"loss": 1.2062,
|
| 247 |
+
"step": 68
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.3482587064676617,
|
| 251 |
+
"grad_norm": 4.614293098449707,
|
| 252 |
+
"learning_rate": 1e-05,
|
| 253 |
+
"loss": 1.7538,
|
| 254 |
+
"step": 70
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.3582089552238806,
|
| 258 |
+
"grad_norm": 6.363886833190918,
|
| 259 |
+
"learning_rate": 1e-05,
|
| 260 |
+
"loss": 1.8623,
|
| 261 |
+
"step": 72
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.3681592039800995,
|
| 265 |
+
"grad_norm": 4.428086757659912,
|
| 266 |
+
"learning_rate": 1e-05,
|
| 267 |
+
"loss": 1.3132,
|
| 268 |
+
"step": 74
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.3781094527363184,
|
| 272 |
+
"grad_norm": 4.011998653411865,
|
| 273 |
+
"learning_rate": 1e-05,
|
| 274 |
+
"loss": 1.4121,
|
| 275 |
+
"step": 76
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.3880597014925373,
|
| 279 |
+
"grad_norm": 4.77931547164917,
|
| 280 |
+
"learning_rate": 1e-05,
|
| 281 |
+
"loss": 1.1911,
|
| 282 |
+
"step": 78
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.39800995024875624,
|
| 286 |
+
"grad_norm": 3.3552842140197754,
|
| 287 |
+
"learning_rate": 1e-05,
|
| 288 |
+
"loss": 1.7105,
|
| 289 |
+
"step": 80
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.4079601990049751,
|
| 293 |
+
"grad_norm": 4.957729339599609,
|
| 294 |
+
"learning_rate": 1e-05,
|
| 295 |
+
"loss": 1.3459,
|
| 296 |
+
"step": 82
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.417910447761194,
|
| 300 |
+
"grad_norm": 9.953269004821777,
|
| 301 |
+
"learning_rate": 1e-05,
|
| 302 |
+
"loss": 1.3948,
|
| 303 |
+
"step": 84
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.42786069651741293,
|
| 307 |
+
"grad_norm": 4.559720039367676,
|
| 308 |
+
"learning_rate": 1e-05,
|
| 309 |
+
"loss": 0.9833,
|
| 310 |
+
"step": 86
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.43781094527363185,
|
| 314 |
+
"grad_norm": 7.114284515380859,
|
| 315 |
+
"learning_rate": 1e-05,
|
| 316 |
+
"loss": 1.6281,
|
| 317 |
+
"step": 88
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.44776119402985076,
|
| 321 |
+
"grad_norm": 1.6471774578094482,
|
| 322 |
+
"learning_rate": 1e-05,
|
| 323 |
+
"loss": 0.5938,
|
| 324 |
+
"step": 90
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.4577114427860697,
|
| 328 |
+
"grad_norm": 6.015200614929199,
|
| 329 |
+
"learning_rate": 1e-05,
|
| 330 |
+
"loss": 0.8881,
|
| 331 |
+
"step": 92
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.46766169154228854,
|
| 335 |
+
"grad_norm": 5.2069807052612305,
|
| 336 |
+
"learning_rate": 1e-05,
|
| 337 |
+
"loss": 0.8571,
|
| 338 |
+
"step": 94
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.47761194029850745,
|
| 342 |
+
"grad_norm": 3.1441423892974854,
|
| 343 |
+
"learning_rate": 1e-05,
|
| 344 |
+
"loss": 1.0354,
|
| 345 |
+
"step": 96
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.48756218905472637,
|
| 349 |
+
"grad_norm": 5.655142307281494,
|
| 350 |
+
"learning_rate": 1e-05,
|
| 351 |
+
"loss": 1.5471,
|
| 352 |
+
"step": 98
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.4975124378109453,
|
| 356 |
+
"grad_norm": 0.610934853553772,
|
| 357 |
+
"learning_rate": 1e-05,
|
| 358 |
+
"loss": 1.0387,
|
| 359 |
+
"step": 100
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.5074626865671642,
|
| 363 |
+
"grad_norm": 3.9553608894348145,
|
| 364 |
+
"learning_rate": 1e-05,
|
| 365 |
+
"loss": 1.5511,
|
| 366 |
+
"step": 102
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5174129353233831,
|
| 370 |
+
"grad_norm": 4.1815643310546875,
|
| 371 |
+
"learning_rate": 1e-05,
|
| 372 |
+
"loss": 2.1176,
|
| 373 |
+
"step": 104
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.527363184079602,
|
| 377 |
+
"grad_norm": 2.519489049911499,
|
| 378 |
+
"learning_rate": 1e-05,
|
| 379 |
+
"loss": 1.154,
|
| 380 |
+
"step": 106
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.5373134328358209,
|
| 384 |
+
"grad_norm": 4.286269664764404,
|
| 385 |
+
"learning_rate": 1e-05,
|
| 386 |
+
"loss": 0.5473,
|
| 387 |
+
"step": 108
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.5472636815920398,
|
| 391 |
+
"grad_norm": 3.5862629413604736,
|
| 392 |
+
"learning_rate": 1e-05,
|
| 393 |
+
"loss": 0.8261,
|
| 394 |
+
"step": 110
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.5572139303482587,
|
| 398 |
+
"grad_norm": 4.422317981719971,
|
| 399 |
+
"learning_rate": 1e-05,
|
| 400 |
+
"loss": 0.5126,
|
| 401 |
+
"step": 112
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.5671641791044776,
|
| 405 |
+
"grad_norm": 5.550352573394775,
|
| 406 |
+
"learning_rate": 1e-05,
|
| 407 |
+
"loss": 0.8303,
|
| 408 |
+
"step": 114
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.5771144278606966,
|
| 412 |
+
"grad_norm": 5.213322639465332,
|
| 413 |
+
"learning_rate": 1e-05,
|
| 414 |
+
"loss": 0.7523,
|
| 415 |
+
"step": 116
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5870646766169154,
|
| 419 |
+
"grad_norm": 11.186556816101074,
|
| 420 |
+
"learning_rate": 1e-05,
|
| 421 |
+
"loss": 2.9817,
|
| 422 |
+
"step": 118
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.5970149253731343,
|
| 426 |
+
"grad_norm": 9.335715293884277,
|
| 427 |
+
"learning_rate": 1e-05,
|
| 428 |
+
"loss": 1.9022,
|
| 429 |
+
"step": 120
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.6069651741293532,
|
| 433 |
+
"grad_norm": 2.7171871662139893,
|
| 434 |
+
"learning_rate": 1e-05,
|
| 435 |
+
"loss": 1.2432,
|
| 436 |
+
"step": 122
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.6169154228855721,
|
| 440 |
+
"grad_norm": 4.649653434753418,
|
| 441 |
+
"learning_rate": 1e-05,
|
| 442 |
+
"loss": 1.0988,
|
| 443 |
+
"step": 124
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.6268656716417911,
|
| 447 |
+
"grad_norm": 3.9890730381011963,
|
| 448 |
+
"learning_rate": 1e-05,
|
| 449 |
+
"loss": 0.5893,
|
| 450 |
+
"step": 126
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.6368159203980099,
|
| 454 |
+
"grad_norm": 6.460565567016602,
|
| 455 |
+
"learning_rate": 1e-05,
|
| 456 |
+
"loss": 0.512,
|
| 457 |
+
"step": 128
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.6467661691542289,
|
| 461 |
+
"grad_norm": 3.9557695388793945,
|
| 462 |
+
"learning_rate": 1e-05,
|
| 463 |
+
"loss": 0.8792,
|
| 464 |
+
"step": 130
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.6567164179104478,
|
| 468 |
+
"grad_norm": 7.4455366134643555,
|
| 469 |
+
"learning_rate": 1e-05,
|
| 470 |
+
"loss": 1.1522,
|
| 471 |
+
"step": 132
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.6666666666666666,
|
| 475 |
+
"grad_norm": 6.534711837768555,
|
| 476 |
+
"learning_rate": 1e-05,
|
| 477 |
+
"loss": 1.9208,
|
| 478 |
+
"step": 134
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.6766169154228856,
|
| 482 |
+
"grad_norm": 6.830433368682861,
|
| 483 |
+
"learning_rate": 1e-05,
|
| 484 |
+
"loss": 0.8374,
|
| 485 |
+
"step": 136
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6865671641791045,
|
| 489 |
+
"grad_norm": 11.555503845214844,
|
| 490 |
+
"learning_rate": 1e-05,
|
| 491 |
+
"loss": 2.4355,
|
| 492 |
+
"step": 138
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.6965174129353234,
|
| 496 |
+
"grad_norm": 5.917487621307373,
|
| 497 |
+
"learning_rate": 1e-05,
|
| 498 |
+
"loss": 1.6808,
|
| 499 |
+
"step": 140
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.7064676616915423,
|
| 503 |
+
"grad_norm": 5.900700569152832,
|
| 504 |
+
"learning_rate": 1e-05,
|
| 505 |
+
"loss": 1.6379,
|
| 506 |
+
"step": 142
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.7164179104477612,
|
| 510 |
+
"grad_norm": 3.194807291030884,
|
| 511 |
+
"learning_rate": 1e-05,
|
| 512 |
+
"loss": 1.4734,
|
| 513 |
+
"step": 144
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.7263681592039801,
|
| 517 |
+
"grad_norm": 3.9584128856658936,
|
| 518 |
+
"learning_rate": 1e-05,
|
| 519 |
+
"loss": 0.8805,
|
| 520 |
+
"step": 146
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.736318407960199,
|
| 524 |
+
"grad_norm": 5.527011394500732,
|
| 525 |
+
"learning_rate": 1e-05,
|
| 526 |
+
"loss": 1.0644,
|
| 527 |
+
"step": 148
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.746268656716418,
|
| 531 |
+
"grad_norm": 0.5414027571678162,
|
| 532 |
+
"learning_rate": 1e-05,
|
| 533 |
+
"loss": 0.3768,
|
| 534 |
+
"step": 150
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.7562189054726368,
|
| 538 |
+
"grad_norm": 5.110873222351074,
|
| 539 |
+
"learning_rate": 1e-05,
|
| 540 |
+
"loss": 1.1494,
|
| 541 |
+
"step": 152
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.7661691542288557,
|
| 545 |
+
"grad_norm": 3.829423666000366,
|
| 546 |
+
"learning_rate": 1e-05,
|
| 547 |
+
"loss": 1.3383,
|
| 548 |
+
"step": 154
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.7761194029850746,
|
| 552 |
+
"grad_norm": 5.544273853302002,
|
| 553 |
+
"learning_rate": 1e-05,
|
| 554 |
+
"loss": 1.0189,
|
| 555 |
+
"step": 156
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.7860696517412935,
|
| 559 |
+
"grad_norm": 4.544299602508545,
|
| 560 |
+
"learning_rate": 1e-05,
|
| 561 |
+
"loss": 0.8228,
|
| 562 |
+
"step": 158
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.7960199004975125,
|
| 566 |
+
"grad_norm": 2.10514760017395,
|
| 567 |
+
"learning_rate": 1e-05,
|
| 568 |
+
"loss": 0.4396,
|
| 569 |
+
"step": 160
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.8059701492537313,
|
| 573 |
+
"grad_norm": 8.759244918823242,
|
| 574 |
+
"learning_rate": 1e-05,
|
| 575 |
+
"loss": 2.6229,
|
| 576 |
+
"step": 162
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.8159203980099502,
|
| 580 |
+
"grad_norm": 4.493800640106201,
|
| 581 |
+
"learning_rate": 1e-05,
|
| 582 |
+
"loss": 1.0382,
|
| 583 |
+
"step": 164
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.8258706467661692,
|
| 587 |
+
"grad_norm": 6.032639026641846,
|
| 588 |
+
"learning_rate": 1e-05,
|
| 589 |
+
"loss": 1.2432,
|
| 590 |
+
"step": 166
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.835820895522388,
|
| 594 |
+
"grad_norm": 4.806446552276611,
|
| 595 |
+
"learning_rate": 1e-05,
|
| 596 |
+
"loss": 0.972,
|
| 597 |
+
"step": 168
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.845771144278607,
|
| 601 |
+
"grad_norm": 2.7649338245391846,
|
| 602 |
+
"learning_rate": 1e-05,
|
| 603 |
+
"loss": 0.82,
|
| 604 |
+
"step": 170
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.8557213930348259,
|
| 608 |
+
"grad_norm": 5.7260236740112305,
|
| 609 |
+
"learning_rate": 1e-05,
|
| 610 |
+
"loss": 0.6877,
|
| 611 |
+
"step": 172
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.8656716417910447,
|
| 615 |
+
"grad_norm": 3.464872121810913,
|
| 616 |
+
"learning_rate": 1e-05,
|
| 617 |
+
"loss": 0.7152,
|
| 618 |
+
"step": 174
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.8756218905472637,
|
| 622 |
+
"grad_norm": 5.897403240203857,
|
| 623 |
+
"learning_rate": 1e-05,
|
| 624 |
+
"loss": 0.5449,
|
| 625 |
+
"step": 176
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.8855721393034826,
|
| 629 |
+
"grad_norm": 6.831718921661377,
|
| 630 |
+
"learning_rate": 1e-05,
|
| 631 |
+
"loss": 0.5211,
|
| 632 |
+
"step": 178
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.8955223880597015,
|
| 636 |
+
"grad_norm": 7.2938103675842285,
|
| 637 |
+
"learning_rate": 1e-05,
|
| 638 |
+
"loss": 1.0614,
|
| 639 |
+
"step": 180
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.9054726368159204,
|
| 643 |
+
"grad_norm": 5.426947116851807,
|
| 644 |
+
"learning_rate": 1e-05,
|
| 645 |
+
"loss": 0.9279,
|
| 646 |
+
"step": 182
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.9154228855721394,
|
| 650 |
+
"grad_norm": 5.0720906257629395,
|
| 651 |
+
"learning_rate": 1e-05,
|
| 652 |
+
"loss": 2.4255,
|
| 653 |
+
"step": 184
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.9253731343283582,
|
| 657 |
+
"grad_norm": 3.3232831954956055,
|
| 658 |
+
"learning_rate": 1e-05,
|
| 659 |
+
"loss": 1.2572,
|
| 660 |
+
"step": 186
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.9353233830845771,
|
| 664 |
+
"grad_norm": 4.725320816040039,
|
| 665 |
+
"learning_rate": 1e-05,
|
| 666 |
+
"loss": 1.7172,
|
| 667 |
+
"step": 188
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.945273631840796,
|
| 671 |
+
"grad_norm": 8.049622535705566,
|
| 672 |
+
"learning_rate": 1e-05,
|
| 673 |
+
"loss": 0.891,
|
| 674 |
+
"step": 190
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.9552238805970149,
|
| 678 |
+
"grad_norm": 3.4230241775512695,
|
| 679 |
+
"learning_rate": 1e-05,
|
| 680 |
+
"loss": 0.6561,
|
| 681 |
+
"step": 192
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.9651741293532339,
|
| 685 |
+
"grad_norm": 4.571541786193848,
|
| 686 |
+
"learning_rate": 1e-05,
|
| 687 |
+
"loss": 1.0751,
|
| 688 |
+
"step": 194
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.9751243781094527,
|
| 692 |
+
"grad_norm": 4.519652366638184,
|
| 693 |
+
"learning_rate": 1e-05,
|
| 694 |
+
"loss": 1.7025,
|
| 695 |
+
"step": 196
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.9850746268656716,
|
| 699 |
+
"grad_norm": 3.518521308898926,
|
| 700 |
+
"learning_rate": 1e-05,
|
| 701 |
+
"loss": 0.3863,
|
| 702 |
+
"step": 198
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.9950248756218906,
|
| 706 |
+
"grad_norm": 0.9762414693832397,
|
| 707 |
+
"learning_rate": 1e-05,
|
| 708 |
+
"loss": 0.0726,
|
| 709 |
+
"step": 200
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.0,
|
| 713 |
+
"step": 201,
|
| 714 |
+
"total_flos": 3.816239406461747e+16,
|
| 715 |
+
"train_loss": 1.273851043549343,
|
| 716 |
+
"train_runtime": 504.9988,
|
| 717 |
+
"train_samples_per_second": 1.592,
|
| 718 |
+
"train_steps_per_second": 0.398
|
| 719 |
+
}
|
| 720 |
+
],
|
| 721 |
+
"logging_steps": 2,
|
| 722 |
+
"max_steps": 201,
|
| 723 |
+
"num_input_tokens_seen": 0,
|
| 724 |
+
"num_train_epochs": 1,
|
| 725 |
+
"save_steps": 500,
|
| 726 |
+
"stateful_callbacks": {
|
| 727 |
+
"TrainerControl": {
|
| 728 |
+
"args": {
|
| 729 |
+
"should_epoch_stop": false,
|
| 730 |
+
"should_evaluate": false,
|
| 731 |
+
"should_log": false,
|
| 732 |
+
"should_save": false,
|
| 733 |
+
"should_training_stop": false
|
| 734 |
+
},
|
| 735 |
+
"attributes": {}
|
| 736 |
+
}
|
| 737 |
+
},
|
| 738 |
+
"total_flos": 3.816239406461747e+16,
|
| 739 |
+
"train_batch_size": 1,
|
| 740 |
+
"trial_name": null,
|
| 741 |
+
"trial_params": null
|
| 742 |
+
}
|