Add models

.ipynb_checkpoints/README-checkpoint.md

@@ -0,0 +1,74 @@
+---
+language:
+  - "en"
+thumbnail: "https://example.com/path/to/your/thumbnail.jpg" # URL to a thumbnail used in social sharing
+tags:
+- "tag1" # For example, "sentiment-analysis"
+- "tag2" # For example, "machine-translation"
+license: "mit"
+datasets:
+- "dataset1" # For example, "imdb"
+- "dataset2" # For example, "wmt16"
+metrics:
+- "metric1" # For example, "accuracy"
+- "metric2" # For example, "f1"
+---
+
+# Your Model Name
+
+## Introduction
+
+This is a brief introduction about your transformer-based model. Here, you can mention the type of the model, the task it was trained for, its performance, and other key features or highlights.
+
+## Training
+
+Here, give detailed information about how the model was trained:
+
+- Dataset(s) used for training
+- Preprocessing techniques used
+- Training configuration such as the batch size, learning rate, optimizer, number of epochs, etc.
+- Any specific challenges or notable aspects of the training process
+
+## Usage
+
+Provide examples of how to use the model for inference. You can provide both a simple usage case and a more complex one if necessary. Make sure to explain what the inputs and outputs are.
+
+Here's a basic example:
+
+from transformers import AutoTokenizer, AutoModel
+
+tokenizer = AutoTokenizer.from_pretrained("your-model-name")
+model = AutoModel.from_pretrained("your-model-name")
+
+inputs = tokenizer("Your example sentence", return_tensors="pt")
+outputs = model(**inputs)
+
+# Explain what the outputs are
+
+## Evaluation
+
+Discuss how the model was evaluated, which metrics were used, and what results it achieved.
+
+## Limitations and Bias
+
+Every model has its limitations and may have certain biases due to the data it was trained on. Explain those here.
+
+## About Us
+
+A small introduction about you or your team. 
+
+## Acknowledgments
+
+Thank people, organizations or mention the resources that helped you in this work.
+
+## License
+
+This model is distributed under the MIT license.
+
+## Contact
+
+Provide a contact method (e.g., email or GitHub issues) for people to reach out with questions, comments, or concerns.
+
+## References
+
+List any relevant references for your model here.

model_M/config.json

@@ -0,0 +1 @@
+{"vocab_size": 24, "query_size": 512, "key_size": 512, "value_size": 512, "num_hiddens": 512, "num_layers": 6, "dropout": 0.2, "lr": 0.0004, "training_steps": 300000, "batch_size": 4096, "label_smoothing": 0.1, "ffn_num_input": 512, "ffn_num_hiddens": 2048, "num_heads": 8, "norm_shape": [512], "device": "cpu"}

model_M/model_weights.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:197a52e0a47711602b2ae60f2f9ee34b15fdc20b195aad66ab8df0bda9aa3256
+size 176577875

model_M_retrain/config.json

@@ -0,0 +1 @@
+{"vocab_size": 24, "query_size": 512, "key_size": 512, "value_size": 512, "num_hiddens": 512, "num_layers": 6, "dropout": 0.2, "lr": 0.0004, "training_steps": 300000, "batch_size": 4096, "label_smoothing": 0.1, "ffn_num_input": 512, "ffn_num_hiddens": 2048, "num_heads": 8, "norm_shape": [512], "device": "cpu"}

model_M_retrain/model_weights.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:596171bfa552fd963cf7e3f1bd8b6c6f35f7ea14d129d9f7b451d0ecb4014ddc
+size 176577875