h2oai / h2o-llmstudio

H2O LLM Studio - a framework and no-code GUI for fine-tuning LLMs. Documentation: https://h2oai.github.io/h2o-llmstudio/

https://gpt-gm.h2o.ai

h2oai/h2o-llmstudio Issues

[FEATURE] add support for multigpu, splitting model across gpus without using deepspeed/fsdp
Updated in 19 hours
[FEATURE] Add DoRA: Weight-Decomposed Low-Rank Adaptation
Updated 5 hours ago1
[BUG] HuggingFace export does not preserve bfloat16 weights but converts to float16 silently when using CPU for upload
Updated 3 days ago9
Data Format section has a broken link
Closed 6 days ago4
[FEATURE] Option to plot train/eval plots with epoch instead of step on x-axis
Updated 7 days ago
[FEATURE] Add UI tests to .github actions
Updated 9 days ago
[FEATURE] Support for minimum learning rate
Updated 9 days ago1
[FEATURE] Use local LLM deployment as Judge
Updated 9 days ago1
Compare Zero-Epoch Prediction with Fine-Tuned Prediction as well as Validaiton Score Comparison
Updated 12 days ago
[FEATURE] Option for not saving checkpoint
Closed 13 days ago1
ValueError: invalid literal for int() with base 10: ‘Failed to initialize NVML: Unknown Error’
Updated 13 days ago8
[FEATURE] Option to add BOS token
Updated 15 days ago4
[FEATURE] Add ability to specify dataset problem type during data import
Updated 19 days ago1
[BUG] Chat window generation parameters not updated
Closed 20 days ago
[FEATURE] Random validation sample for chat interface
Updated 21 days ago
[CODE IMPROVEMENT] Use build-in H2O Wave Chatbot features
Closed 21 days ago
[BUG] UI freezes when using "Stop streaming" button with a text in the input box
Closed 21 days ago
[FEATURE] Fine-tune CohereForCausalLM Models
Closed 21 days ago
[BUG] Exporting / downloading model larger, than VRAM available (trained with DeepSpeed) fails
Updated 22 days ago8
Inference model on distributed GPU after model training via CLI
Closed 22 days ago3
[DOC] Update the code examples to demonstrate how to implement experiments from CLI using YAML configuration files
Closed 22 days ago
[CODE IMPROVEMENT] Default for max_time
Closed 23 days ago
[FEATURE] Mixed Precision Dtype
Closed 23 days ago
[BUG] Mixed precision not working with bfloat16
Closed 23 days ago2
[CODE IMPROVEMENT] Avoid gradient sync during gradient accumulation
Closed 23 days ago
[CODE IMPROVEMENT] Sort data files alphabetically
Closed 23 days ago
CUDA out of memory
Closed 24 days ago6
[BUG] Pipenv missing as a requirement for the `make llmstudio` command
Closed a month ago1
[CHORE] Upgrade dependencies to fix vulnerabilities
Closed a month ago
[FEATURE] Add additional digit of precision for specifying learning rate and other parameters
Closed a month ago1
[FEATURE] Add danube2 to default model list
Closed a month ago
[BUG] Scheduler should consider gradient accumulation while assigning `epoch_steps`?
Updated a month ago
[CODE IMPROVEMENT] Flash_attn installation may be wrong if the wheel is cached
Closed a month ago
[FEATURE] Use Free Form Chat Template to Define Input Styling
Updated a month ago
[BUG] Tokenizer config has add_bos_token=true while LLM Studio is training with add_special_tokens=False
Updated 2 months ago1
[BUG] Downloading a fine-tuned model fails with error
Closed 2 months ago2
[DOCS] Adding the description syntax
Closed 2 months ago
[BUG] Error when pushing model to HuggingFace
Closed 2 months ago2
[CODE IMPROVEMENT] Resolve High Vulnerabilities
Closed 2 months ago1
[FEATURE] Training with QLoRA + FSDP
Updated 2 months ago
[BUG] Recent regression causing an error when loading tokenizer for Deepseek models
Closed 2 months ago2
Can you kindly provide documentation for macOS
Closed 3 months ago1
[BUG] Unable to access LLM studio using Docker from a nightly build
Closed 3 months ago2
[CODE IMPROVEMENT] Create CICD for UI Testing
Closed 3 months ago1
[FEATURE]create distributed_inference.sh with deepspeed
Closed 3 months ago3
[FEATURE] MoE Aux Loss
Updated 3 months ago
[CODE IMPROVEMENT] if user specifies a “system” column but it doesnt exist, it should error out instead of continue running silently
Updated 3 months ago5
[BUG] Default dataset can not be written to the database
Closed 3 months ago
[BUG] download model zip does not contain safetensors
Closed 3 months ago
[FEATURE]when saving multiple epochs add an epoch number suffix for when save best=False
Updated 4 months ago3