Update README.md

README.md

@@ -1,17 +1,18 @@
 ---
 base_model:
-- ethicalabs/xLSTM-7b-Instruct-PEFT
+- NX-AI/xLSTM-7b
 library_name: transformers
 model_name: xlstm-7b-chatml
 tags:
-- base_model:adapter:NX-AI/xLSTM-7b
-- lora
+- base_model:adapter:ethicalabs/xLSTM-7b-Instruct-PEFT
 - sft
 - transformers
 - trl
 licence: license
 pipeline_tag: text-generation
 license: mit
+datasets:
+- HuggingFaceH4/ultrachat_200k
 ---
 
 # Model Card for xLSTM-7b-Instruct
@@ -21,15 +22,56 @@ It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start
 
+For text generation you have to to pin specific pytorch versions [https://huggingface.co/datasets/John6666/forum1/blob/main/xlstm_1.md](https://huggingface.co/datasets/John6666/forum1/blob/main/xlstm_1.md)
+
+```shell
+pip install "torch==2.5.1" "torchvision==0.20.1" "torchaudio==2.5.1" --index-url https://download.pytorch.org/whl/cu124
+pip install "triton==3.4.0"  # >=3.1 is OK; 3.4.0 current as of Sep 2025
+pip install "mlstm-kernels==2.0.1" "xlstm==2.0.5"
+```
+
 ```python
-TODO
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer, AutoConfig
+
+MERGED_MODEL_PATH = "ethicalabs/xLSTM-7b-Instruct"
+
+# We apply a configuration that uses native, hardware-agnostic kernels.
+print("Defining a safe, native kernel configuration for compatibility...")
+
+safe_config = AutoConfig.from_pretrained(MERGED_MODEL_PATH, trust_remote_code=True)
+# Use the stable, native parallel kernel
+safe_config.chunkwise_kernel = "chunkwise--native_autograd"
+safe_config.sequence_kernel = "native_sequence__native"
+safe_config.step_kernel = "native"
+# This flag is still required for the HF implementation to avoid unpacking errors
+safe_config.return_last_states = False
+
+# Load the final, merged model with the safe config (no quantization)
+print("Loading the final, merged model in bfloat16 (no quantization for compatibility)...")
+final_model = AutoModelForCausalLM.from_pretrained(
+    MERGED_MODEL_PATH,
+    device_map="auto",
+    torch_dtype=torch.bfloat16,
+    trust_remote_code=True,
+    config=safe_config 
+)
+final_tokenizer = AutoTokenizer.from_pretrained(MERGED_MODEL_PATH)
+
+# The tokenizer needs to know which token to use for padding.
+if final_tokenizer.pad_token is None:
+    final_tokenizer.pad_token = final_tokenizer.eos_token
+print("Padding token has been set.")
+
+
+# Set the model to evaluation mode
+final_model.eval()
 ```
 
 ## Training procedure
 
 [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/ethicalabs-ai/xlstm-finetuning/runs/pfmf34a3) 
 
-
 This model was trained with SFT.
 
 ### Framework versions
@@ -43,8 +85,6 @@ This model was trained with SFT.
 
 ## Citations
 
-
-
 Cite TRL as:
     
 ```bibtex