Spaces:

K446
/

Opengrid

Sleeping

App Files Files Community

Opengrid

Commit History

Apr 26

11:10
Delete generate_plots.py

K446

Verified signature
1d191e4
10:52
docs(readme): point blog links to the HF Space copy of blog.md⁠

K446

7b306cb
10:46
docs: clarify scenario count, OPENGRID_MODE flag; drop runtime/epoch info⁠

K446

2f2ff77
10:29
Polish for hackathon submission: training evidence, two pipelines, UI, docs⁠

K446

e81353d
03:11
Fix health check timeout: start UI server in background before training

K446

89992e4
01:01
Print every reward call so terminal shows continuous progress⁠

K446

81257d9
00:51
Reduce prompt/completion length to fix silent OOM on backward pass

K446

a76abcc
00:49
Replace env-simulation reward with fast pure-heuristic to fix hang⁠

K446

efbeb4b
00:02
Fix batch/num_generations: 4/4 on 1 GPU, grad_accum=4

K446

114859b

Apr 25

23:53
Trigger HF Space rebuild for batch_size=4 fix⁠

K446

d50432b
23:45
Fix GRPO training: reward variance, batch/gen alignment, generation config⁠

K446

e1ab78c
23:40
Update run_training.py and train_grpo.py, remove Dockerfile.training

K446

7be88b4
23:26
Add pre-train gen sanity check, explicit GenerationConfig, dynamic GRPOConfig params, torch_compile/vllm off

K446

a6ecb81
23:18
QLoRA best practices: prepare_model_for_kbit_training, paged_adamw_8bit, cosine LR, faster iteration

K446

8dab919
23:03
Fix: enable_input_require_grads for gradient checkpointing + 4-bit

K446

c505237
22:57
Fix OOM: reduce batch/gen/tokens, add grad checkpointing + adafactor

K446

c09f4cb
22:48
Fix batch size: 8 to match num_generations=8

K446

8d94e97
22:42
Drop unsloth: use standard bitsandbytes 4-bit + peft LoRA + TRL GRPOTrainer

K446

6072ace
22:27
Pin transformers <4.52 and unsloth_zoo==2025.11.1 for API compat

K446

b724812
22:22
Pin TRL <0.16 for unsloth 2025.11.1 GRPO API compatibility

K446

08c4515
22:17
Dynamic NVIDIA lib path discovery in entrypoint for bitsandbytes

K446

c7e8b79
22:10
Add LD_LIBRARY_PATH for pip-installed NVIDIA libs (bitsandbytes fix)

K446

f9c90fc
22:04
Add torchvision and hf_transfer

K446

00b8117
21:56
Remove torchao entirely - transformers handles absence gracefully

K446

d4e5470
21:51
Fix: install torchao 0.8.0 separately, unsloth --no-deps to avoid torchao>=0.13 conflict

K446

f4d773c
21:44
Pin compatible versions: torch 2.6.0 + torchao <0.9 + transformers <5.0

K446

9b70933
21:36
Update CUDA to 12.4.1 and unpin PyTorch version to fix torchao/int1 compatibility

K446

371b620
21:30
Remove --no-deps to allow installing sub-dependencies like regex

K446

3c0ad6e
21:16
Add explicit compiler location debugging in run_training.py

K446

1991472
21:04
Explicitly install gcc, g++, python3-dev and set CC/CXX

K446

8bac226
20:59
Add build-essential for Triton compilation

K446

2b3d2c3
20:41
GRPO training with CUDA + results in UI

K446

bcce6af
19:22
fix: notebook uses compute_grpo_reward_env, updated hyperparams, no emojis

K446

69bab30
18:52
fix: remove ENV OPENGRID_MODE to avoid HF secrets collision

K446

be15396
18:50
fix: transformers>=4.51.3 to resolve unsloth dep conflict

K446

3313fd3
18:48
fix: add unsloth back with pinned versions to avoid dep backtracking

K446

689cb35
18:46
fix: lean Dockerfile + remove unsloth from training deps

K446

b2a04c7
18:40
fix: unified Dockerfile with entrypoint for server/training mode⁠

K446

1dfed79
18:33
fix: update .dockerignore for new repo structure

K446

db3026a
18:27
feat: curriculum training + Karnataka scenarios + repo cleanup⁠

K446

8a02303
16:35
restore env Dockerfile

K446

5ba53f6
16:34
training dockerfile

K446

05e16e8
16:34
fix: total_memory attribute name

K446

8cdf625
16:23
Restore env Dockerfile

K446

7299ddc
16:20
Use training Dockerfile for GPU Space

K446

af0b804
16:19
Add GRPO training runner for HF Spaces GPU

K446

7bcd08c

Apr 6

23:01
OpenGrid: Multi-agent POMDP power grid environment with GRPO training⁠

K446

78131a0

Commit History

Apr 26

Delete generate_plots.py

docs(readme): point blog links to the HF Space copy of blog.md⁠

docs: clarify scenario count, OPENGRID_MODE flag; drop runtime/epoch info⁠

Polish for hackathon submission: training evidence, two pipelines, UI, docs⁠

Fix health check timeout: start UI server in background before training

Print every reward call so terminal shows continuous progress⁠

Reduce prompt/completion length to fix silent OOM on backward pass

Replace env-simulation reward with fast pure-heuristic to fix hang⁠

Fix batch/num_generations: 4/4 on 1 GPU, grad_accum=4

Apr 25

Trigger HF Space rebuild for batch_size=4 fix⁠

Fix GRPO training: reward variance, batch/gen alignment, generation config⁠

Update run_training.py and train_grpo.py, remove Dockerfile.training

Add pre-train gen sanity check, explicit GenerationConfig, dynamic GRPOConfig params, torch_compile/vllm off

QLoRA best practices: prepare_model_for_kbit_training, paged_adamw_8bit, cosine LR, faster iteration

Fix: enable_input_require_grads for gradient checkpointing + 4-bit

Fix OOM: reduce batch/gen/tokens, add grad checkpointing + adafactor

Fix batch size: 8 to match num_generations=8

Drop unsloth: use standard bitsandbytes 4-bit + peft LoRA + TRL GRPOTrainer

Pin transformers <4.52 and unsloth_zoo==2025.11.1 for API compat

Pin TRL <0.16 for unsloth 2025.11.1 GRPO API compatibility

Dynamic NVIDIA lib path discovery in entrypoint for bitsandbytes

Add LD_LIBRARY_PATH for pip-installed NVIDIA libs (bitsandbytes fix)

Add torchvision and hf_transfer

Remove torchao entirely - transformers handles absence gracefully

Fix: install torchao 0.8.0 separately, unsloth --no-deps to avoid torchao>=0.13 conflict

Pin compatible versions: torch 2.6.0 + torchao <0.9 + transformers <5.0

Update CUDA to 12.4.1 and unpin PyTorch version to fix torchao/int1 compatibility

Remove --no-deps to allow installing sub-dependencies like regex

Add explicit compiler location debugging in run_training.py

Explicitly install gcc, g++, python3-dev and set CC/CXX

Add build-essential for Triton compilation

GRPO training with CUDA + results in UI

fix: notebook uses compute_grpo_reward_env, updated hyperparams, no emojis

fix: remove ENV OPENGRID_MODE to avoid HF secrets collision

fix: transformers>=4.51.3 to resolve unsloth dep conflict

fix: add unsloth back with pinned versions to avoid dep backtracking

fix: lean Dockerfile + remove unsloth from training deps

fix: unified Dockerfile with entrypoint for server/training mode⁠

fix: update .dockerignore for new repo structure

feat: curriculum training + Karnataka scenarios + repo cleanup⁠

restore env Dockerfile

training dockerfile

fix: total_memory attribute name

Restore env Dockerfile

Use training Dockerfile for GPU Space

Add GRPO training runner for HF Spaces GPU

Apr 6

OpenGrid: Multi-agent POMDP power grid environment with GRPO training⁠