Upload EarthMind-4B GRPO fine-tuned model

README.md

@@ -0,0 +1,130 @@
+---
+license: apache-2.0
+language:
+- en
+tags:
+- vision-language
+- vlm
+- grpo
+- earthmind
+- geospatial
+- remote-sensing
+library_name: transformers
+pipeline_tag: image-text-to-text
+---
+
+# EarthMind-R1
+
+EarthMind-R1 is a vision-language model fine-tuned using GRPO (Group Relative Policy Optimization) for geospatial and remote sensing image understanding tasks.
+
+## Model Description
+
+- **Base Model:** EarthMind-4B
+- **Training Method:** GRPO (Group Relative Policy Optimization)
+- **Training Data:** Geospatial instruction dataset
+- **Fine-tuning:** LoRA adapters merged into base weights
+
+## Usage
+
+### Quick Start
+
+```python
+import torch
+from PIL import Image
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+# Load model and tokenizer
+model_id = "aadex/Earthmind-R1"
+
+tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)
+model = AutoModelForCausalLM.from_pretrained(
+    model_id,
+    trust_remote_code=True,
+    torch_dtype=torch.bfloat16,
+    device_map="auto",
+)
+
+# Load an image
+image = Image.open("your_image.jpg").convert("RGB")
+
+# Ask a question
+question = "Describe what you see in this satellite image."
+
+# Use model's chat interface
+response = model.chat(
+    tokenizer=tokenizer,
+    question=question,
+    images=[image],
+    generation_config={
+        "max_new_tokens": 512,
+        "temperature": 0.7,
+        "do_sample": True,
+    },
+)
+
+print(response)
+```
+
+### Expected Output Format
+
+The model is trained to provide structured responses:
+
+```
+<think>
+[Reasoning about the image content]
+</think>
+<answer>
+[Final answer to the question]
+</answer>
+```
+
+## Requirements
+
+```
+torch>=2.0
+transformers>=4.40
+accelerate
+pillow
+```
+
+## Hardware Requirements
+
+- **Minimum:** 16GB VRAM (with bfloat16)
+- **Recommended:** 24GB VRAM for comfortable inference
+
+## Training Details
+
+- **Framework:** VLM-R1 + TRL
+- **Optimizer:** AdamW
+- **Learning Rate:** 1e-6
+- **LoRA Configuration:**
+  - r: 32
+  - alpha: 64
+  - dropout: 0.05
+- **GRPO Settings:**
+  - num_generations: 4
+  - num_iterations: 2
+  - beta: 0.01
+
+## Limitations
+
+- Optimized for geospatial/remote sensing imagery
+- May not perform as well on general domain images
+- Response quality depends on image resolution and clarity
+
+## Citation
+
+If you use this model, please cite:
+
+```bibtex
+@misc{earthmind-r1,
+  title={EarthMind-R1: GRPO Fine-tuned Vision-Language Model for Geospatial Understanding},
+  author={Your Name},
+  year={2024},
+  publisher={HuggingFace}
+}
+```
+
+## License
+
+Apache 2.0

model-00001-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8c6e07df93c166ba98e65cd235559e512d79b5a92aa3adb1d967c4d1f3a741d4
+oid sha256:97f3792a0d86308d529a858ac40fb0d704ffa3a4da4a042a6acb77b184e5eb97
 size 4993044040

model-00002-of-00002.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c305d389e82ee348d9dbbe3b8fd23637842a98663562bfa94a33aa74bc4554e2
+oid sha256:8a5ada102da6c3ed05981f12a81dc425a3aa173c9e18778530ff3fab08ee9313
 size 2890805372

modeling_earthmind_chat.py

@@ -38,7 +38,9 @@ from types import MethodType
 import torch.nn.functional as F
 
 try:
-    from .flash_attention import FlashAttention
+    # flash_attention import removed for inference without flash_attn
+# from .flash_attention import FlashAttention
+FlashAttention = None
     has_flash_attn = True
 except:
     print('FlashAttention is not installed.')

modeling_intern_vit.py

@@ -21,7 +21,9 @@ from transformers.utils import logging
 from .configuration_intern_vit import InternVisionConfig
 
 try:
-    from .flash_attention import FlashAttention
+    # flash_attention import removed for inference without flash_attn
+# from .flash_attention import FlashAttention
+FlashAttention = None
     has_flash_attn = True
 except:
     print('FlashAttention is not installed.')