Update README.md

README.md

@@ -2,34 +2,61 @@
 license: apache-2.0
 base_model: google/bigbird-roberta-base
 tags:
-- generated_from_trainer
-model-index:
-- name: bigbird-roberta-base-fineweb-edu-llama3-annotations-4096-vN
-  results: []
+- eduscore
+- data filter
+inference: false
+datasets:
+- HuggingFaceFW/fineweb-edu-llama3-annotations
+language:
+- en
 ---
 
-<!-- This model card has been generated automatically according to the information the Trainer had access to. You
-should probably proofread and complete it, then remove this comment. -->
 
 [<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/pszemraj/eduscore-regression/runs/04oc07hx)
-# bigbird-roberta-base-fineweb-edu-llama3-annotations-4096-vN
+# bigbird-roberta-base: eduscore 
+
+Similar to the [original](https://hf.co/HuggingFaceFW/fineweb-edu-classifier), this model predicts a score of 0 to 5 on 'educational quality' of some text. This model was fine-tuned @ its max context length of 4096 tokens.
+
+
+## Usage
+
+Note this is for CPU, for GPU you will need to make some (small) changes.
+
+```py
+# Load model directly
+from transformers import AutoTokenizer, AutoModelForSequenceClassification
+
+model_name = "pszemraj/bigbird-roberta-base-edu-classifier"
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForSequenceClassification.from_pretrained(
+    model_name, attn_implementation="eager"
+)
+
+text = "This is a test sentence."
+inputs = tokenizer(text, return_tensors="pt", padding="longest", truncation=True)
+outputs = model(**inputs)
+logits = outputs.logits.squeeze(-1).float().detach().numpy()
+score = logits.item()
+result = {
+    "text": text,
+    "score": score,
+    "int_score": int(round(max(0, min(score, 5)))),
+}
+
+print(result)
+# {'text': 'This is a test sentence.', 'score': 0.20170727372169495, 'int_score': 0}
+```
+
+## Details
 
 This model is a fine-tuned version of [google/bigbird-roberta-base](https://huggingface.co/google/bigbird-roberta-base) on the HuggingFaceFW/fineweb-edu-llama3-annotations dataset.
 It achieves the following results on the evaluation set:
 - Loss: 0.2176
 - Mse: 0.2176
 
-## Model description
-
-More information needed
-
 ## Intended uses & limitations
 
-More information needed
-
-## Training and evaluation data
-
-More information needed
+Refer to the hf classifier's [model card](https://huggingface.co/HuggingFaceFW/fineweb-edu-classifier#limitations) for more details
 
 ## Training procedure
 
@@ -45,51 +72,4 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.98) and epsilon=1e-09
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_ratio: 0.05
-- num_epochs: 1.0
-
-### Training results
-
-| Training Loss | Epoch  | Step | Validation Loss | Mse    |
-|:-------------:|:------:|:----:|:---------------:|:------:|
-| 0.4763        | 0.0288 | 100  | 0.4468          | 0.4468 |
-| 0.3078        | 0.0577 | 200  | 0.3130          | 0.3130 |
-| 0.3088        | 0.0865 | 300  | 0.2695          | 0.2695 |
-| 0.2379        | 0.1153 | 400  | 0.2618          | 0.2618 |
-| 0.289         | 0.1441 | 500  | 0.2583          | 0.2583 |
-| 0.3049        | 0.1730 | 600  | 0.2723          | 0.2723 |
-| 0.2292        | 0.2018 | 700  | 0.2477          | 0.2477 |
-| 0.2677        | 0.2306 | 800  | 0.2369          | 0.2369 |
-| 0.3181        | 0.2594 | 900  | 0.2307          | 0.2307 |
-| 0.2551        | 0.2883 | 1000 | 0.2411          | 0.2411 |
-| 0.2743        | 0.3171 | 1100 | 0.2350          | 0.2350 |
-| 0.2383        | 0.3459 | 1200 | 0.2424          | 0.2424 |
-| 0.2191        | 0.3747 | 1300 | 0.2279          | 0.2279 |
-| 0.2431        | 0.4036 | 1400 | 0.2232          | 0.2232 |
-| 0.2161        | 0.4324 | 1500 | 0.2307          | 0.2307 |
-| 0.2459        | 0.4612 | 1600 | 0.2246          | 0.2246 |
-| 0.2403        | 0.4900 | 1700 | 0.2232          | 0.2232 |
-| 0.251         | 0.5189 | 1800 | 0.2421          | 0.2421 |
-| 0.2565        | 0.5477 | 1900 | 0.2207          | 0.2207 |
-| 0.2274        | 0.5765 | 2000 | 0.2294          | 0.2294 |
-| 0.2272        | 0.6053 | 2100 | 0.2192          | 0.2192 |
-| 0.2668        | 0.6342 | 2200 | 0.2204          | 0.2204 |
-| 0.2434        | 0.6630 | 2300 | 0.2196          | 0.2196 |
-| 0.2464        | 0.6918 | 2400 | 0.2185          | 0.2185 |
-| 0.2338        | 0.7206 | 2500 | 0.2166          | 0.2166 |
-| 0.243         | 0.7495 | 2600 | 0.2165          | 0.2165 |
-| 0.1891        | 0.7783 | 2700 | 0.2201          | 0.2201 |
-| 0.2355        | 0.8071 | 2800 | 0.2167          | 0.2167 |
-| 0.2231        | 0.8359 | 2900 | 0.2168          | 0.2168 |
-| 0.2274        | 0.8648 | 3000 | 0.2243          | 0.2243 |
-| 0.2287        | 0.8936 | 3100 | 0.2203          | 0.2203 |
-| 0.261         | 0.9224 | 3200 | 0.2186          | 0.2186 |
-| 0.2187        | 0.9512 | 3300 | 0.2176          | 0.2176 |
-| 0.2069        | 0.9801 | 3400 | 0.2178          | 0.2178 |
-
-
-### Framework versions
-
-- Transformers 4.42.3
-- Pytorch 2.3.1+cu121
-- Datasets 2.20.0
-- Tokenizers 0.19.1
+- num_epochs: 1.0