Image-Text-to-Text
PEFT
Safetensors
GGUF
microscopy
vision-language
diatoms
fungal-spores
biology
bioindicator
gemma-4
unsloth
qlora
multimodal
on-device
offline
conversational
Instructions to use Laborator/microlens-final with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use Laborator/microlens-final with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("unsloth/gemma-4-e2b-it-unsloth-bnb-4bit") model = PeftModel.from_pretrained(base_model, "Laborator/microlens-final") - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- llama.cpp
How to use Laborator/microlens-final with llama.cpp:
Install (macOS, Linux)
curl -LsSf https://llama.app/install.sh | sh # Start a local OpenAI-compatible server with a web UI: llama serve -hf Laborator/microlens-final # Run inference directly in the terminal: llama cli -hf Laborator/microlens-final
Install from WinGet (Windows)
winget install llama.cpp # Start a local OpenAI-compatible server with a web UI: llama serve -hf Laborator/microlens-final # Run inference directly in the terminal: llama cli -hf Laborator/microlens-final
Build from source code
git clone https://github.com/ggml-org/llama.cpp.git cd llama.cpp cmake -B build cmake --build build -j --target llama-app # Start a local OpenAI-compatible server with a web UI: ./build/bin/llama serve -hf Laborator/microlens-final # Run inference directly in the terminal: ./build/bin/llama cli -hf Laborator/microlens-final
Use Docker
docker model run hf.co/Laborator/microlens-final
- Llama new
- LM Studio
- Jan
- vLLM
How to use Laborator/microlens-final with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "Laborator/microlens-final" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "Laborator/microlens-final", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "Describe this image in one sentence." }, { "type": "image_url", "image_url": { "url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg" } } ] } ] }'Use Docker
docker model run hf.co/Laborator/microlens-final
- Ollama
How to use Laborator/microlens-final with Ollama:
ollama run hf.co/Laborator/microlens-final
- Unsloth Desktop
- Pi
How to use Laborator/microlens-final with Pi:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf Laborator/microlens-final
Configure the model in Pi
# Install Pi: npm install -g @earendil-works/pi-coding-agent # Add to ~/.pi/agent/models.json: { "providers": { "llama-cpp": { "baseUrl": "http://localhost:9931/v1", "api": "openai-completions", "apiKey": "none", "models": [ { "id": "Laborator/microlens-final" } ] } } }Run Pi
# Start Pi in your project directory: pi
- Docker Model Runner
How to use Laborator/microlens-final with Docker Model Runner:
docker model run hf.co/Laborator/microlens-final
- Lemonade
How to use Laborator/microlens-final with Lemonade:
Pull the model
# Download Lemonade from https://lemonade-server.ai/ lemonade pull Laborator/microlens-final
Run and chat with the model
lemonade run user.microlens-final-{{QUANT_TAG}}List all available models
lemonade list
- Hermes Agent
How to use Laborator/microlens-final with Hermes Agent:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf Laborator/microlens-final
Configure Hermes
# Install Hermes: curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash hermes setup # Point Hermes at the local server: hermes config set model.provider custom hermes config set model.base_url http://127.0.0.1:9931/v1 hermes config set model.default Laborator/microlens-final
Run Hermes
hermes
- Atomic Chat
- OpenClaw
How to use Laborator/microlens-final with OpenClaw:
Start the llama.cpp server
# Install llama.cpp: brew install llama.cpp # Start a local OpenAI-compatible server: llama serve -hf Laborator/microlens-final
Configure OpenClaw
# Install OpenClaw: npm install -g openclaw@latest # Register the local server and set it as the default model: openclaw onboard --non-interactive --mode local \ --auth-choice custom-api-key \ --custom-base-url http://127.0.0.1:9931/v1 \ --custom-model-id "Laborator/microlens-final" \ --custom-provider-id llama-cpp \ --custom-compatibility openai \ --custom-text-input \ --accept-risk \ --skip-health
Run OpenClaw
openclaw agent --local --agent main --message "Hello from Hugging Face"
docs: replace auto-generated card with full MicroLens card (hackathon submission)
Browse files
README.md
CHANGED
|
@@ -1,210 +1,165 @@
|
|
| 1 |
---
|
| 2 |
-
|
|
|
|
| 3 |
library_name: peft
|
| 4 |
-
pipeline_tag: text-
|
| 5 |
tags:
|
| 6 |
-
-
|
| 7 |
-
-
|
| 8 |
-
-
|
| 9 |
-
-
|
| 10 |
-
-
|
|
|
|
|
|
|
| 11 |
- unsloth
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
---
|
| 13 |
|
| 14 |
-
#
|
| 15 |
|
| 16 |
-
|
| 17 |
|
|
|
|
| 18 |
|
|
|
|
| 19 |
|
| 20 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
|
| 22 |
-
##
|
| 23 |
|
| 24 |
-
|
| 25 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
|
|
|
|
| 27 |
|
| 28 |
-
|
| 29 |
-
- **Funded by [optional]:** [More Information Needed]
|
| 30 |
-
- **Shared by [optional]:** [More Information Needed]
|
| 31 |
-
- **Model type:** [More Information Needed]
|
| 32 |
-
- **Language(s) (NLP):** [More Information Needed]
|
| 33 |
-
- **License:** [More Information Needed]
|
| 34 |
-
- **Finetuned from model [optional]:** [More Information Needed]
|
| 35 |
|
| 36 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 37 |
|
| 38 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 43 |
|
| 44 |
-
##
|
| 45 |
|
| 46 |
-
|
|
|
|
|
|
|
| 47 |
|
| 48 |
-
|
| 49 |
|
| 50 |
-
|
| 51 |
|
| 52 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 53 |
|
| 54 |
-
##
|
| 55 |
|
| 56 |
-
|
| 57 |
|
| 58 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 59 |
|
| 60 |
-
|
| 61 |
|
| 62 |
-
|
| 63 |
|
| 64 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
|
| 66 |
-
|
| 67 |
|
| 68 |
-
|
| 69 |
|
| 70 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 71 |
|
| 72 |
-
##
|
| 73 |
|
| 74 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 75 |
|
| 76 |
-
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
[More Information Needed]
|
| 83 |
-
|
| 84 |
-
## Training Details
|
| 85 |
-
|
| 86 |
-
### Training Data
|
| 87 |
-
|
| 88 |
-
<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
|
| 89 |
-
|
| 90 |
-
[More Information Needed]
|
| 91 |
-
|
| 92 |
-
### Training Procedure
|
| 93 |
-
|
| 94 |
-
<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
|
| 95 |
-
|
| 96 |
-
#### Preprocessing [optional]
|
| 97 |
-
|
| 98 |
-
[More Information Needed]
|
| 99 |
-
|
| 100 |
-
|
| 101 |
-
#### Training Hyperparameters
|
| 102 |
-
|
| 103 |
-
- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
| 104 |
-
|
| 105 |
-
#### Speeds, Sizes, Times [optional]
|
| 106 |
-
|
| 107 |
-
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
| 108 |
-
|
| 109 |
-
[More Information Needed]
|
| 110 |
-
|
| 111 |
-
## Evaluation
|
| 112 |
-
|
| 113 |
-
<!-- This section describes the evaluation protocols and provides the results. -->
|
| 114 |
-
|
| 115 |
-
### Testing Data, Factors & Metrics
|
| 116 |
-
|
| 117 |
-
#### Testing Data
|
| 118 |
-
|
| 119 |
-
<!-- This should link to a Dataset Card if possible. -->
|
| 120 |
-
|
| 121 |
-
[More Information Needed]
|
| 122 |
-
|
| 123 |
-
#### Factors
|
| 124 |
-
|
| 125 |
-
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 126 |
-
|
| 127 |
-
[More Information Needed]
|
| 128 |
-
|
| 129 |
-
#### Metrics
|
| 130 |
-
|
| 131 |
-
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 132 |
-
|
| 133 |
-
[More Information Needed]
|
| 134 |
-
|
| 135 |
-
### Results
|
| 136 |
-
|
| 137 |
-
[More Information Needed]
|
| 138 |
-
|
| 139 |
-
#### Summary
|
| 140 |
-
|
| 141 |
-
|
| 142 |
-
|
| 143 |
-
## Model Examination [optional]
|
| 144 |
-
|
| 145 |
-
<!-- Relevant interpretability work for the model goes here -->
|
| 146 |
-
|
| 147 |
-
[More Information Needed]
|
| 148 |
-
|
| 149 |
-
## Environmental Impact
|
| 150 |
-
|
| 151 |
-
<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
|
| 152 |
-
|
| 153 |
-
Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
|
| 154 |
-
|
| 155 |
-
- **Hardware Type:** [More Information Needed]
|
| 156 |
-
- **Hours used:** [More Information Needed]
|
| 157 |
-
- **Cloud Provider:** [More Information Needed]
|
| 158 |
-
- **Compute Region:** [More Information Needed]
|
| 159 |
-
- **Carbon Emitted:** [More Information Needed]
|
| 160 |
-
|
| 161 |
-
## Technical Specifications [optional]
|
| 162 |
-
|
| 163 |
-
### Model Architecture and Objective
|
| 164 |
-
|
| 165 |
-
[More Information Needed]
|
| 166 |
-
|
| 167 |
-
### Compute Infrastructure
|
| 168 |
-
|
| 169 |
-
[More Information Needed]
|
| 170 |
-
|
| 171 |
-
#### Hardware
|
| 172 |
-
|
| 173 |
-
[More Information Needed]
|
| 174 |
-
|
| 175 |
-
#### Software
|
| 176 |
-
|
| 177 |
-
[More Information Needed]
|
| 178 |
-
|
| 179 |
-
## Citation [optional]
|
| 180 |
-
|
| 181 |
-
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
| 182 |
-
|
| 183 |
-
**BibTeX:**
|
| 184 |
-
|
| 185 |
-
[More Information Needed]
|
| 186 |
-
|
| 187 |
-
**APA:**
|
| 188 |
-
|
| 189 |
-
[More Information Needed]
|
| 190 |
-
|
| 191 |
-
## Glossary [optional]
|
| 192 |
-
|
| 193 |
-
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
| 194 |
-
|
| 195 |
-
[More Information Needed]
|
| 196 |
-
|
| 197 |
-
## More Information [optional]
|
| 198 |
-
|
| 199 |
-
[More Information Needed]
|
| 200 |
-
|
| 201 |
-
## Model Card Authors [optional]
|
| 202 |
-
|
| 203 |
-
[More Information Needed]
|
| 204 |
-
|
| 205 |
-
## Model Card Contact
|
| 206 |
-
|
| 207 |
-
[More Information Needed]
|
| 208 |
-
### Framework versions
|
| 209 |
-
|
| 210 |
-
- PEFT 0.18.0
|
|
|
|
| 1 |
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
base_model: unsloth/gemma-4-E2B-it
|
| 4 |
library_name: peft
|
| 5 |
+
pipeline_tag: image-text-to-text
|
| 6 |
tags:
|
| 7 |
+
- microscopy
|
| 8 |
+
- vision-language
|
| 9 |
+
- diatoms
|
| 10 |
+
- fungal-spores
|
| 11 |
+
- biology
|
| 12 |
+
- bioindicator
|
| 13 |
+
- gemma-4
|
| 14 |
- unsloth
|
| 15 |
+
- qlora
|
| 16 |
+
- multimodal
|
| 17 |
+
- on-device
|
| 18 |
+
- offline
|
| 19 |
+
datasets:
|
| 20 |
+
- sergheibrinza/microlens-vqa-hackathon
|
| 21 |
+
- sergheibrinza/microlens-images-hackathon
|
| 22 |
+
language:
|
| 23 |
+
- en
|
| 24 |
+
- de
|
| 25 |
+
- fr
|
| 26 |
+
- es
|
| 27 |
+
- it
|
| 28 |
+
- pt
|
| 29 |
+
- ru
|
| 30 |
+
- zh
|
| 31 |
+
- ja
|
| 32 |
+
- ko
|
| 33 |
---
|
| 34 |
|
| 35 |
+
# MicroLens — Final
|
| 36 |
|
| 37 |
+
**A pocket-microscope expert.** Vision-language model that identifies microscopy specimens — diatoms and fungal spores across 95 genera — names the genus, and explains morphology, habitat, and identification cues. Built on Gemma 4 E2B, runs offline on a 4 GB Android, speaks 140+ languages out of the box.
|
| 38 |
|
| 39 |
+
Submission to the **Kaggle Gemma 4 Good Hackathon 2026**.
|
| 40 |
|
| 41 |
+
## Links
|
| 42 |
|
| 43 |
+
| Resource | URL |
|
| 44 |
+
|---|---|
|
| 45 |
+
| Live web demo | https://huggingface.co/spaces/Laborator/microlens |
|
| 46 |
+
| Live Kaggle notebook (T4, 9 min) | https://www.kaggle.com/code/sergheibrinza/microlens-final |
|
| 47 |
+
| GitHub (source, APK, Modelfile) | https://github.com/SergheiBrinza/microlens |
|
| 48 |
+
| Training VQA dataset (75,491 pairs) | https://www.kaggle.com/datasets/sergheibrinza/microlens-vqa-hackathon |
|
| 49 |
+
| Training images (75,491 PNGs) | https://www.kaggle.com/datasets/sergheibrinza/microlens-images-hackathon |
|
| 50 |
+
| Ollama (3 GB GGUF) | `ollama run brinzaengineeringai/microlens-final` |
|
| 51 |
+
| Android APK | https://github.com/SergheiBrinza/microlens/releases |
|
| 52 |
|
| 53 |
+
## What this model is
|
| 54 |
|
| 55 |
+
A 4-bit QLoRA fine-tune of `unsloth/gemma-4-E2B-it` that turns a generic vision-language model into a structured microscopy assistant. For any specimen image, MicroLens returns four sections:
|
| 56 |
|
| 57 |
+
- **Genus** (and species when it is sure)
|
| 58 |
+
- **Morphology** — shape, size, raphe, frustule
|
| 59 |
+
- **Habitat** — where this organism typically lives
|
| 60 |
+
- **Identification cues** — what to look for in the image
|
| 61 |
|
| 62 |
+
Covers **95 genera** across two categories: diatoms (the standard bioindicator behind the EU Water Framework Directive) and fungal spores.
|
| 63 |
|
| 64 |
+
## Quick start (Python + Unsloth)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 65 |
|
| 66 |
+
```python
|
| 67 |
+
from unsloth import FastVisionModel
|
| 68 |
+
from peft import PeftModel
|
| 69 |
+
from PIL import Image
|
| 70 |
+
import torch
|
| 71 |
|
| 72 |
+
base, tokenizer = FastVisionModel.from_pretrained(
|
| 73 |
+
'unsloth/gemma-4-E2B-it',
|
| 74 |
+
load_in_4bit=True,
|
| 75 |
+
use_gradient_checkpointing='unsloth',
|
| 76 |
+
max_seq_length=2048,
|
| 77 |
+
)
|
| 78 |
+
model = PeftModel.from_pretrained(base, 'Laborator/microlens-final')
|
| 79 |
+
FastVisionModel.for_inference(model)
|
| 80 |
|
| 81 |
+
img = Image.open('your_specimen.png').convert('RGB')
|
| 82 |
+
prompt = 'Identify the organism in this microscopy image and describe its morphology.'
|
| 83 |
+
msgs = [{'role':'user','content':[{'type':'image'},{'type':'text','text':prompt}]}]
|
| 84 |
+
text = tokenizer.apply_chat_template(msgs, add_generation_prompt=True)
|
| 85 |
+
inp = tokenizer(img, text, add_special_tokens=False, return_tensors='pt').to('cuda')
|
| 86 |
+
out = model.generate(**inp, max_new_tokens=200, do_sample=False)
|
| 87 |
+
print(tokenizer.decode(out[0][inp.input_ids.shape[-1]:], skip_special_tokens=True))
|
| 88 |
+
```
|
| 89 |
|
| 90 |
+
## Quick start (Ollama, on-device)
|
| 91 |
|
| 92 |
+
```bash
|
| 93 |
+
ollama run brinzaengineeringai/microlens-final
|
| 94 |
+
```
|
| 95 |
|
| 96 |
+
Pulls the 3 GB Q4_K_M GGUF and runs entirely on CPU or any consumer GPU.
|
| 97 |
|
| 98 |
+
## Training summary
|
| 99 |
|
| 100 |
+
- **Base model:** `unsloth/gemma-4-E2B-it` (4.44 B parameters, ~2 B effective via Per-Layer Embeddings)
|
| 101 |
+
- **Method:** 4-bit QLoRA via Unsloth FastVisionModel, both vision tower and language tower trainable
|
| 102 |
+
- **Data:** 75,491 VQA pairs (67,121 train + 8,370 val), 95 genera, 2 categories
|
| 103 |
+
- **Schedule:** 2 epochs, 8,392 steps, lr 2e-4 cosine, batch 2×8=16, AdamW-8bit, bf16, seq 2048
|
| 104 |
+
- **Hardware:** 1× RTX 3090 Ti (24 GB), 14.7 hours wall-clock
|
| 105 |
+
- **Trainable params:** 29.9 M (0.58% of base), LoRA r=16, α=32
|
| 106 |
+
- **Final eval loss:** 0.0189 (smooth monotone decrease)
|
| 107 |
|
| 108 |
+
## Evaluation results
|
| 109 |
|
| 110 |
+
Stratified 200-pair validation, 150 diatom + 50 fungal spore.
|
| 111 |
|
| 112 |
+
| Metric | Diatom (n=150) | Fungal spore (n=50) | Overall (n=200) |
|
| 113 |
+
|---|---|---|---|
|
| 114 |
+
| **Genus accuracy** (substring match) | 85.3% | **100%** | **89.0%** |
|
| 115 |
+
| **Category accuracy** | 100% | 100% | **100%** |
|
| 116 |
+
| **Format adherence** (morphology + habitat + cues) | 95.3% | 72.0% | **89.5%** |
|
| 117 |
|
| 118 |
+
Reproducible end to end on a free Kaggle T4 in 9 minutes — see the linked Kaggle notebook.
|
| 119 |
|
| 120 |
+
## Training data — license-clean for commercial use
|
| 121 |
|
| 122 |
+
| Source | License | Pairs (train) |
|
| 123 |
+
|---|---|---|
|
| 124 |
+
| UDE Diatoms in the Wild 2024 (Zenodo 10410655) | CC0 | 39,389 |
|
| 125 |
+
| DIATLAS (Zenodo 16260887) | CC-BY 4.0 | 23,544 |
|
| 126 |
+
| TgFC — Tectona grandis fungal community (figshare 28855910) | CC-BY 4.0 | 4,188 |
|
| 127 |
|
| 128 |
+
Top-30 genera have hand-curated knowledge-base answers from AlgaeBase, WoRMS, ITIS. ZooScanNet (CC-BY-NC) and ZooLake (license not verifiable) were intentionally excluded so this release is clean for commercial use end to end.
|
| 129 |
|
| 130 |
+
## Honest limits
|
| 131 |
|
| 132 |
+
- Trained on stained light-microscopy at 384×384. SEM and fluorescence are out of distribution.
|
| 133 |
+
- Only 95 genera (diatoms + fungal spores). Plankton, pollen and soil microbes were in earlier internal builds but got dropped to keep licensing clean.
|
| 134 |
+
- Long-tail genera produce shorter answers. The curated knowledge base only covers the top 30.
|
| 135 |
+
- Confidence is expressed in words ("looks like X but the asymmetry suggests Y"), not calibrated probabilities. Good for an explainable assistant, bad for automated decisions.
|
| 136 |
+
- No held-out test split. The 8,370 val pairs do double duty for per-step and final eval. A future release will fix that.
|
| 137 |
+
- **Research artefact — not a medical device. Not for clinical, diagnostic, or regulatory use.**
|
| 138 |
|
| 139 |
+
## License & attribution
|
| 140 |
|
| 141 |
+
Apache 2.0 — matches base Gemma 4 license. Please credit *Serghei Brinza — MicroLens, Vienna 2026*.
|
| 142 |
+
|
| 143 |
+
## Citation
|
| 144 |
+
|
| 145 |
+
If you use MicroLens in research, please cite:
|
| 146 |
+
|
| 147 |
+
```bibtex
|
| 148 |
+
@misc{brinza2026microlens,
|
| 149 |
+
author = {Serghei Brinza},
|
| 150 |
+
title = {MicroLens: A Pocket-Microscope Expert via Gemma 4 E2B},
|
| 151 |
+
year = 2026,
|
| 152 |
+
publisher = {Hugging Face},
|
| 153 |
+
howpublished = {\url{https://huggingface.co/Laborator/microlens-final}},
|
| 154 |
+
note = {Kaggle Gemma 4 Good Hackathon 2026 submission}
|
| 155 |
+
}
|
| 156 |
+
```
|
| 157 |
+
|
| 158 |
+
Also cite the upstream:
|
| 159 |
|
| 160 |
+
- Gemma 4 (Google DeepMind)
|
| 161 |
+
- Unsloth (Daniel & Michael Han) — https://github.com/unslothai/unsloth
|
| 162 |
+
- AlgaeBase, WoRMS, ITIS — taxonomic knowledge bases
|
| 163 |
+
- UDE Diatoms in the Wild 2024 (Zenodo 10410655)
|
| 164 |
+
- DIATLAS (Zenodo 16260887)
|
| 165 |
+
- TgFC (figshare 28855910)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|