Save model at global step 260
Browse files- config.yaml +5 -5
- data.pt +1 -1
- extra_state_world_size_4_rank_0.pt +1 -1
- extra_state_world_size_4_rank_1.pt +1 -1
- extra_state_world_size_4_rank_2.pt +1 -1
- extra_state_world_size_4_rank_3.pt +1 -1
- model_world_size_4_rank_0.pt +1 -1
- model_world_size_4_rank_1.pt +1 -1
- model_world_size_4_rank_2.pt +1 -1
- model_world_size_4_rank_3.pt +1 -1
- optim_world_size_4_rank_0.pt +1 -1
- optim_world_size_4_rank_1.pt +1 -1
- optim_world_size_4_rank_2.pt +1 -1
- optim_world_size_4_rank_3.pt +1 -1
config.yaml
CHANGED
@@ -61,7 +61,7 @@ actor_rollout_ref:
|
|
61 |
lr_warmup_steps_ratio: 0.0
|
62 |
min_lr_ratio: null
|
63 |
warmup_style: constant
|
64 |
-
total_training_steps:
|
65 |
weight_decay: 0.01
|
66 |
fsdp_config:
|
67 |
wrap_policy:
|
@@ -122,7 +122,7 @@ critic:
|
|
122 |
lr_warmup_steps_ratio: 0.0
|
123 |
min_lr_ratio: null
|
124 |
warmup_style: constant
|
125 |
-
total_training_steps:
|
126 |
weight_decay: 0.01
|
127 |
model:
|
128 |
path: ~/models/deepseek-llm-7b-chat
|
@@ -192,8 +192,8 @@ algorithm:
|
|
192 |
target_kl: 0.1
|
193 |
trainer:
|
194 |
balance_batch: true
|
195 |
-
total_epochs:
|
196 |
-
total_training_steps:
|
197 |
project_name: value-LLM
|
198 |
experiment_name: grpo-aime24-qwen2.5math-1.5B-base-mbs128-n4
|
199 |
logger:
|
@@ -218,7 +218,7 @@ trainer:
|
|
218 |
resume_from_hf:
|
219 |
enable: true
|
220 |
actor_hf_repo_id: RyanYr/grpo-aime24-qwen2.5math-1.5B-base-mbs128-n4_actor
|
221 |
-
actor_revision:
|
222 |
critic_hf_repo_id: null
|
223 |
critic_revision: main
|
224 |
hf_token: null
|
|
|
61 |
lr_warmup_steps_ratio: 0.0
|
62 |
min_lr_ratio: null
|
63 |
warmup_style: constant
|
64 |
+
total_training_steps: 1000
|
65 |
weight_decay: 0.01
|
66 |
fsdp_config:
|
67 |
wrap_policy:
|
|
|
122 |
lr_warmup_steps_ratio: 0.0
|
123 |
min_lr_ratio: null
|
124 |
warmup_style: constant
|
125 |
+
total_training_steps: 1000
|
126 |
weight_decay: 0.01
|
127 |
model:
|
128 |
path: ~/models/deepseek-llm-7b-chat
|
|
|
192 |
target_kl: 0.1
|
193 |
trainer:
|
194 |
balance_batch: true
|
195 |
+
total_epochs: 1000000000000
|
196 |
+
total_training_steps: 1000
|
197 |
project_name: value-LLM
|
198 |
experiment_name: grpo-aime24-qwen2.5math-1.5B-base-mbs128-n4
|
199 |
logger:
|
|
|
218 |
resume_from_hf:
|
219 |
enable: true
|
220 |
actor_hf_repo_id: RyanYr/grpo-aime24-qwen2.5math-1.5B-base-mbs128-n4_actor
|
221 |
+
actor_revision: 0eda92b4d077e223d32c4e146810e9c20eabf30e
|
222 |
critic_hf_repo_id: null
|
223 |
critic_revision: main
|
224 |
hf_token: null
|
data.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1492
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:6bf5116bec0225580dbdbc6b00e299d696095e1964ae7b4a01ab4a50fd0afd94
|
3 |
size 1492
|
extra_state_world_size_4_rank_0.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14632
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f238b9d88f95c43ce0358c6aa29d3a8ac83bfde38e7de9499c1ef5947753735c
|
3 |
size 14632
|
extra_state_world_size_4_rank_1.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14632
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:95459f83480a478f0413f89f74cdad33637dccde1a417b6f0b5e175c15e9215d
|
3 |
size 14632
|
extra_state_world_size_4_rank_2.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14632
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8a02ca18deca1aa76cbd2268e352342240caa42c2f85a2d868f7f53cae20bb6f
|
3 |
size 14632
|
extra_state_world_size_4_rank_3.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 14632
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a0ef7398fb6a8729f2653a6cc78edd94298b9e19882b69348953fa85511e8f9f
|
3 |
size 14632
|
model_world_size_4_rank_0.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1777276538
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:a48d3c9f272d29c6aa59faae90bf73bef7a12fbd82a59d1a132ac1af99f80828
|
3 |
size 1777276538
|
model_world_size_4_rank_1.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1777276538
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:088bd34027d70c52b6db86ccf92ca3efcda2715002e4db3474300531cabe3eaa
|
3 |
size 1777276538
|
model_world_size_4_rank_2.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1777276538
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:59e02e721d8b92f0a71e2a4036995a86f2f75d8576c8160b89ac28e6aa374879
|
3 |
size 1777276538
|
model_world_size_4_rank_3.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 1777276538
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7ca7100f3189d32c9884bc62ff67d359ba091cde46cea837ee65eef9cb7fd36a
|
3 |
size 1777276538
|
optim_world_size_4_rank_0.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3087454775
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:12696c9898e3176f84005a66099f560323e38a43ff93679b18a2c2220b65872b
|
3 |
size 3087454775
|
optim_world_size_4_rank_1.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3087454775
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:62cfc208cb55774e908dc5f7d7d5973510de41cf37f2e186f9fcf5c71244ef15
|
3 |
size 3087454775
|
optim_world_size_4_rank_2.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3087454775
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:df38044ec6724f0155ef6da2b5ad5749a06e92449ea2a854c08aac9f3edf535e
|
3 |
size 3087454775
|
optim_world_size_4_rank_3.pt
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 3087454775
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ae70e75512c088bf2c5ff53f9247e5b9e1e779225cbcdf7c555b10db71284e8e
|
3 |
size 3087454775
|