eddierex commited on
Commit
ddde4dd
·
verified ·
1 Parent(s): 08f6817

Upload GR00T_N1_5 checkpoint 20000 for OpenWBC bottle task

Browse files
experiment_cfg/conf.yaml CHANGED
@@ -65,7 +65,7 @@ model:
65
  data:
66
  datasets:
67
  - dataset_paths:
68
- - /home/opuser/wtx/vla/data/new/onlybottom_gr00t
69
  embodiment_tag: new_embodiment
70
  mix_ratio: 1.0
71
  dataset_type: physical_embodiment
@@ -187,9 +187,9 @@ data:
187
  - 224
188
  video_backend: torchcodec
189
  training:
190
- output_dir: ./save/test1
191
  experiment_name: null
192
- max_steps: 20000
193
  global_batch_size: 96
194
  batch_size: null
195
  gradient_accumulation_steps: 1
@@ -200,13 +200,13 @@ training:
200
  warmup_steps: 0
201
  max_grad_norm: 1.0
202
  optim: adamw_torch
203
- start_from_checkpoint: ../models/GR00T-N1.6-G1-PnPAppleToPlate/
204
  tf32: true
205
  fp16: false
206
  bf16: true
207
  eval_bf16: true
208
  logging_steps: 10
209
- save_steps: 1000
210
  save_total_limit: 5
211
  save_vl_model: false
212
  upload_checkpoints: false
@@ -241,5 +241,5 @@ training:
241
  - 0
242
  open_loop_eval_steps_per_traj: 100
243
  open_loop_eval_plot_indices: null
244
- max_steps: 20000
245
- save_steps: 1000
 
65
  data:
66
  datasets:
67
  - dataset_paths:
68
+ - /home/opuser/wtx/vla/data/gr00t/new/onlybottom_gr00t
69
  embodiment_tag: new_embodiment
70
  mix_ratio: 1.0
71
  dataset_type: physical_embodiment
 
187
  - 224
188
  video_backend: torchcodec
189
  training:
190
+ output_dir: ./save/test6000
191
  experiment_name: null
192
+ max_steps: 60000
193
  global_batch_size: 96
194
  batch_size: null
195
  gradient_accumulation_steps: 1
 
200
  warmup_steps: 0
201
  max_grad_norm: 1.0
202
  optim: adamw_torch
203
+ start_from_checkpoint: /home/opuser/wtx/vla/models/GR00T-N1.6-G1-PnPAppleToPlate
204
  tf32: true
205
  fp16: false
206
  bf16: true
207
  eval_bf16: true
208
  logging_steps: 10
209
+ save_steps: 5000
210
  save_total_limit: 5
211
  save_vl_model: false
212
  upload_checkpoints: false
 
241
  - 0
242
  open_loop_eval_steps_per_traj: 100
243
  open_loop_eval_plot_indices: null
244
+ max_steps: 60000
245
+ save_steps: 5000
experiment_cfg/config.yaml CHANGED
@@ -4,7 +4,7 @@ data: !!python/object:gr00t.configs.data.data_config.DataConfig
4
  datasets:
5
  - !!python/object:gr00t.configs.data.data_config.SingleDatasetConfig
6
  dataset_paths:
7
- - /home/opuser/wtx/vla/data/new/onlybottom_gr00t
8
  dataset_type: physical_embodiment
9
  embodiment_tag: new_embodiment
10
  mix_ratio: 1.0
@@ -253,21 +253,21 @@ training: !!python/object:gr00t.configs.training.training_config.TrainingConfig
253
  max_concurrent_uploads: 2
254
  max_grad_norm: 1.0
255
  max_retries: 3
256
- max_steps: 20000
257
  num_gpus: 2
258
  open_loop_eval_plot_indices: null
259
  open_loop_eval_steps_per_traj: 100
260
  open_loop_eval_traj_ids:
261
  - 0
262
  optim: adamw_torch
263
- output_dir: ./save/test1
264
  remove_unused_columns: false
265
  save_best_eval_metric_greater_is_better: true
266
  save_best_eval_metric_name: ''
267
- save_steps: 1000
268
  save_total_limit: 5
269
  save_vl_model: false
270
- start_from_checkpoint: ../models/GR00T-N1.6-G1-PnPAppleToPlate/
271
  tf32: true
272
  transformers_access_token: null
273
  transformers_cache_dir: null
 
4
  datasets:
5
  - !!python/object:gr00t.configs.data.data_config.SingleDatasetConfig
6
  dataset_paths:
7
+ - /home/opuser/wtx/vla/data/gr00t/new/onlybottom_gr00t
8
  dataset_type: physical_embodiment
9
  embodiment_tag: new_embodiment
10
  mix_ratio: 1.0
 
253
  max_concurrent_uploads: 2
254
  max_grad_norm: 1.0
255
  max_retries: 3
256
+ max_steps: 60000
257
  num_gpus: 2
258
  open_loop_eval_plot_indices: null
259
  open_loop_eval_steps_per_traj: 100
260
  open_loop_eval_traj_ids:
261
  - 0
262
  optim: adamw_torch
263
+ output_dir: ./save/test6000
264
  remove_unused_columns: false
265
  save_best_eval_metric_greater_is_better: true
266
  save_best_eval_metric_name: ''
267
+ save_steps: 5000
268
  save_total_limit: 5
269
  save_vl_model: false
270
+ start_from_checkpoint: /home/opuser/wtx/vla/models/GR00T-N1.6-G1-PnPAppleToPlate
271
  tf32: true
272
  transformers_access_token: null
273
  transformers_cache_dir: null
global_step60000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:63d230f64928175c6af4f17e0412783538ab366382fa6058eaa6bdcd6dcf3c5c
3
+ size 9719841825
global_step60000/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef845a54f74aca7e7d730957939008f393d7db61e49a03e3c44b4cdb33bae64a
3
+ size 9719835873
global_step60000/mp_rank_00_model_states.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b0981877d4d94ba5df862920ecd3eb841288114e4888b7fea1581331af8fb103
3
+ size 9907202435
latest CHANGED
@@ -1 +1 @@
1
- global_step20000
 
1
+ global_step60000
model-00001-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a229a9f6dedeaee028cd1b60f4f8e1f938836efb98d39b422f4c5ad8c665b02b
3
  size 4991091456
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf5bae92cdb4e52d89fb76282f84cb0c1aa811ef4252c6261bcf3a8ae5b71e9a
3
  size 4991091456
model-00002-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:af1879144a86a457d2603f333003a55a59e9b0d5906143828728afcf75d6380f
3
  size 1582283096
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5153e1d8be1ea2726732c057906ddcf3309f6ac5a0236d2a6b839696b5b1975
3
  size 1582283096
rng_state_0.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3a2d386a33a3a3e726b5af06a1adef33f3da8e695771b1410c8be86d10f29b3b
3
  size 14917
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a1258e46605de576a62034dd10ba29c0a34e9a8238cf43f02aedbc426594eac
3
  size 14917
rng_state_1.pth CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e4d3c79f81b9d18e67f3cc1809d5633c90ae75a4916c495e3f9cb6a8f5483ace
3
  size 14917
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26b5993f4cffa03a121b1ee1670a40a77f949b7a4e76e5102ab72e91290e8f59
3
  size 14917
scheduler.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7fa802a80def971b73ec74284a6aa44d0b2ea101bd38ed41a3b1c1a0b4001f00
3
  size 1465
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:96b418fbb509f96d95ad52dd99d48eeb782a306175bb55d2661eb9b3585613ba
3
  size 1465
trainer_state.json CHANGED
The diff for this file is too large to render. See raw diff
 
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6270b2f9a0a7d10b725ab54b19d53553b72c7707af49e677cf873b583dc2b4d1
3
  size 7633
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:560a23fc414112b4ca651369d3d4ebef45c1454b966cb7683543d996709746be
3
  size 7633
wandb_config.json CHANGED
@@ -1 +1 @@
1
- {"project": "finetune-gr00t-n1d6", "run_id": "test1"}
 
1
+ {"project": "finetune-gr00t-n1d6", "run_id": "test6000"}