forum

1495 threads · Page 4 of 30

Can't fine tune Phi 4 using GRPO 18 messages
Solved
Qwen3 Coder Next on 2x RTX 3090 8 messages
GLM-4.7-Flash not starting on vLLM 3 messages
Qwen3-Coder-Next on 4x5060ti OOM 6 messages
Open
gpt-oss RL takes 41H in DGX rather than 4H as given in docs 2 messages
Open
install unsloth on colab 3 messages
Solved
Unsloth Quant Request: Step 3.5 Flash 11 messages
How do I finetune QLoRA on direct Windows installation? 20 messages
Open
How to inference fine-tuned GPT-OSS-20B saved in mxfp4? 4 messages
Open
How to enable reasoning_content for Kimi-k2.5? 5 messages
Set non-thinking mode during training Qwen3 GRPO 3 messages
Fine-tuning takes very long on H200 and 1B params 143 messages
Solved
Docker image for B200? 4 messages
Data collection framework best practices? 4 messages
Suggested dataset format for full-FT 3 messages
Open
qwen-image-2512 comfy-ui vs stable-diffusion.cpp 2 messages
Open
Multi turn environments like OpenSpiel using RL 29 messages
How to train vision model with IterableDataset? 7 messages
Solved
Ollama glm 4.7 flash reap gguf 6 messages
[Unsloth Docker] Can't import 'FastSentenceTransformer' 22 messages
Cannot install unsloth 25 messages
GLM-4.7 degraded Quality 3 messages
GLM-4.7-FLASH WEIRDNESS IN OLLAMA 7 messages
Open
Unsloth GLM-4.7 Flash Finetuning & RL Request 8 messages
GRPO on A100 80GB GPU resulting in 0.02 it/s. 2 messages
Issue with Multi-GPU Setup for RL Task Using GRPO 7 messages
Is it possible to run GLM-4.7-Flash on 8GB VRAM + 32GB RAM locally or on a Colab notebook 16GB VRAM? 13 messages
Open
Unsloth Quantization Request: GLM-4.7 Flash 2 messages
Extraction with Fine-tuned LLM 2 messages
grad_norm: 0 and LR: 0 17 messages
80gb VRAM usage for Gemma 3 4b 32k Context / Max Sequence 9 messages
Open
Can't get Unsloth to work 22 messages
Support for transformers v5 2 messages
TRL custom rollout func support 38 messages
Enterprise plan 5 messages
Endless thinking with unsloth/Qwen3-Next-80B-A3B-Thinking-GGUF 3 messages
How do I only upload the LORA adapters to huggingface using Unsloth? 34 messages
Nemotron UD 2 XL and UD 3 XL are the same size? 2 messages
Can't create conda environment for unsloth because of xformers [SOLVED] 59 messages
Unsloth Quantization Request: GLM-4.7-REAP (from Cerebras) 2 messages
nan loss: FT Qwen3 QAT 4 messages
No config file found - are you sure the `model_name` is correct? 2 messages
Open
Fine tuning nemotron 3 nano with load_in_4bit doesn't work 12 messages
CISPO / SAPO Loss Support for GRPO? 7 messages
How can I cleanly fine tune unsloth/Ministral-3-3B-Instruct-2512 without vision ? 13 messages
rubbish response from unsloth/Qwen3-Next-80B-A3B-Instruct-GGUF:UD-Q4_K_XL 5 messages
unable to covert Qwen4b-instruct to GGUF 6 messages
Open
RLHF need help chosing correct methods 8 messages
Open
unsloth/Qwen3-VL-30B-A3B-Instruct quantization issue 141 messages
GRPO gpt-oss-20b 100k context OOM & TRL/vLLM Version Hell 239 messages