Commit ·
def8b5b
1
Parent(s): 70b91ba
Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -24,6 +24,29 @@ get_started_code: "``` python \n # Use a pipeline as a high-level helper\n
|
|
| 24 |
\ model = AutoModelForSequenceClassification.from_pretrained(\"MattStammers/test-trainer\"\
|
| 25 |
)\n "
|
| 26 |
training_data: 'see Glue dataset: https://huggingface.co/datasets/glue'
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
model-index:
|
| 28 |
- name: test_trainer
|
| 29 |
results:
|
|
@@ -132,18 +155,18 @@ see Glue dataset: https://huggingface.co/datasets/glue
|
|
| 132 |
|
| 133 |
#### Preprocessing [optional]
|
| 134 |
|
| 135 |
-
|
| 136 |
|
| 137 |
|
| 138 |
#### Training Hyperparameters
|
| 139 |
|
| 140 |
-
- **Training regime:**
|
| 141 |
|
| 142 |
#### Speeds, Sizes, Times [optional]
|
| 143 |
|
| 144 |
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
| 145 |
|
| 146 |
-
|
| 147 |
|
| 148 |
## Evaluation
|
| 149 |
|
|
@@ -155,33 +178,33 @@ see Glue dataset: https://huggingface.co/datasets/glue
|
|
| 155 |
|
| 156 |
<!-- This should link to a Data Card if possible. -->
|
| 157 |
|
| 158 |
-
|
| 159 |
|
| 160 |
#### Factors
|
| 161 |
|
| 162 |
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 163 |
|
| 164 |
-
|
| 165 |
|
| 166 |
#### Metrics
|
| 167 |
|
| 168 |
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 169 |
|
| 170 |
-
|
| 171 |
|
| 172 |
### Results
|
| 173 |
|
| 174 |
-
|
| 175 |
|
| 176 |
#### Summary
|
| 177 |
|
| 178 |
-
|
| 179 |
|
| 180 |
## Model Examination [optional]
|
| 181 |
|
| 182 |
<!-- Relevant interpretability work for the model goes here -->
|
| 183 |
|
| 184 |
-
|
| 185 |
|
| 186 |
## Environmental Impact
|
| 187 |
|
|
@@ -189,29 +212,29 @@ see Glue dataset: https://huggingface.co/datasets/glue
|
|
| 189 |
|
| 190 |
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).
|
| 191 |
|
| 192 |
-
- **Hardware Type:**
|
| 193 |
-
- **Hours used:**
|
| 194 |
-
- **Cloud Provider:**
|
| 195 |
-
- **Compute Region:**
|
| 196 |
-
- **Carbon Emitted:**
|
| 197 |
|
| 198 |
## Technical Specifications [optional]
|
| 199 |
|
| 200 |
### Model Architecture and Objective
|
| 201 |
|
| 202 |
-
|
| 203 |
|
| 204 |
### Compute Infrastructure
|
| 205 |
|
| 206 |
-
|
| 207 |
|
| 208 |
#### Hardware
|
| 209 |
|
| 210 |
-
|
| 211 |
|
| 212 |
#### Software
|
| 213 |
|
| 214 |
-
|
| 215 |
|
| 216 |
## Citation [optional]
|
| 217 |
|
|
@@ -219,28 +242,28 @@ Carbon emissions can be estimated using the [Machine Learning Impact calculator]
|
|
| 219 |
|
| 220 |
**BibTeX:**
|
| 221 |
|
| 222 |
-
|
| 223 |
|
| 224 |
**APA:**
|
| 225 |
|
| 226 |
-
|
| 227 |
|
| 228 |
## Glossary [optional]
|
| 229 |
|
| 230 |
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
| 231 |
|
| 232 |
-
|
| 233 |
|
| 234 |
## More Information [optional]
|
| 235 |
|
| 236 |
-
|
| 237 |
|
| 238 |
## Model Card Authors [optional]
|
| 239 |
|
| 240 |
-
|
| 241 |
|
| 242 |
## Model Card Contact
|
| 243 |
|
| 244 |
-
|
| 245 |
|
| 246 |
|
|
|
|
| 24 |
\ model = AutoModelForSequenceClassification.from_pretrained(\"MattStammers/test-trainer\"\
|
| 25 |
)\n "
|
| 26 |
training_data: 'see Glue dataset: https://huggingface.co/datasets/glue'
|
| 27 |
+
preprocessing: sentence pairs to analyse similarity
|
| 28 |
+
training_regime: user_defined
|
| 29 |
+
speeds_sizes_times: not_relevant
|
| 30 |
+
testing_data: 'MRCP. Link: https://huggingface.co/datasets/SetFit/mrpc'
|
| 31 |
+
testing_factors: unknown
|
| 32 |
+
testing_metrics: unknown
|
| 33 |
+
results: see evaluation results.
|
| 34 |
+
results_summary: results are not bad
|
| 35 |
+
model_examination: model should be interpreted with user discretion.
|
| 36 |
+
model_specs: bert fine-tuned
|
| 37 |
+
compute_infrastructure: requires less than 4GB of GPU to run quickly
|
| 38 |
+
hardware: T600
|
| 39 |
+
hours_used: '0.1'
|
| 40 |
+
cloud_provider: on-prem
|
| 41 |
+
cloud_region: europe
|
| 42 |
+
co2_emitted: very little
|
| 43 |
+
software: python, pytorch with transformers
|
| 44 |
+
citation_bibtex: nil_presently
|
| 45 |
+
citation_apa: not_relevant
|
| 46 |
+
glossary: nil presently
|
| 47 |
+
more_information: can be made available on request
|
| 48 |
+
model_card_authors: Matt Stammers
|
| 49 |
+
model_card_contact: Matt Stammers
|
| 50 |
model-index:
|
| 51 |
- name: test_trainer
|
| 52 |
results:
|
|
|
|
| 155 |
|
| 156 |
#### Preprocessing [optional]
|
| 157 |
|
| 158 |
+
sentence pairs to analyse similarity
|
| 159 |
|
| 160 |
|
| 161 |
#### Training Hyperparameters
|
| 162 |
|
| 163 |
+
- **Training regime:** user_defined <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
|
| 164 |
|
| 165 |
#### Speeds, Sizes, Times [optional]
|
| 166 |
|
| 167 |
<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
|
| 168 |
|
| 169 |
+
not_relevant
|
| 170 |
|
| 171 |
## Evaluation
|
| 172 |
|
|
|
|
| 178 |
|
| 179 |
<!-- This should link to a Data Card if possible. -->
|
| 180 |
|
| 181 |
+
MRCP. Link: https://huggingface.co/datasets/SetFit/mrpc
|
| 182 |
|
| 183 |
#### Factors
|
| 184 |
|
| 185 |
<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
|
| 186 |
|
| 187 |
+
unknown
|
| 188 |
|
| 189 |
#### Metrics
|
| 190 |
|
| 191 |
<!-- These are the evaluation metrics being used, ideally with a description of why. -->
|
| 192 |
|
| 193 |
+
unknown
|
| 194 |
|
| 195 |
### Results
|
| 196 |
|
| 197 |
+
see evaluation results.
|
| 198 |
|
| 199 |
#### Summary
|
| 200 |
|
| 201 |
+
results are not bad
|
| 202 |
|
| 203 |
## Model Examination [optional]
|
| 204 |
|
| 205 |
<!-- Relevant interpretability work for the model goes here -->
|
| 206 |
|
| 207 |
+
model should be interpreted with user discretion.
|
| 208 |
|
| 209 |
## Environmental Impact
|
| 210 |
|
|
|
|
| 212 |
|
| 213 |
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).
|
| 214 |
|
| 215 |
+
- **Hardware Type:** T600
|
| 216 |
+
- **Hours used:** 0.1
|
| 217 |
+
- **Cloud Provider:** on-prem
|
| 218 |
+
- **Compute Region:** europe
|
| 219 |
+
- **Carbon Emitted:** very little
|
| 220 |
|
| 221 |
## Technical Specifications [optional]
|
| 222 |
|
| 223 |
### Model Architecture and Objective
|
| 224 |
|
| 225 |
+
bert fine-tuned
|
| 226 |
|
| 227 |
### Compute Infrastructure
|
| 228 |
|
| 229 |
+
requires less than 4GB of GPU to run quickly
|
| 230 |
|
| 231 |
#### Hardware
|
| 232 |
|
| 233 |
+
T600
|
| 234 |
|
| 235 |
#### Software
|
| 236 |
|
| 237 |
+
python, pytorch with transformers
|
| 238 |
|
| 239 |
## Citation [optional]
|
| 240 |
|
|
|
|
| 242 |
|
| 243 |
**BibTeX:**
|
| 244 |
|
| 245 |
+
nil_presently
|
| 246 |
|
| 247 |
**APA:**
|
| 248 |
|
| 249 |
+
not_relevant
|
| 250 |
|
| 251 |
## Glossary [optional]
|
| 252 |
|
| 253 |
<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
|
| 254 |
|
| 255 |
+
nil presently
|
| 256 |
|
| 257 |
## More Information [optional]
|
| 258 |
|
| 259 |
+
can be made available on request
|
| 260 |
|
| 261 |
## Model Card Authors [optional]
|
| 262 |
|
| 263 |
+
Matt Stammers
|
| 264 |
|
| 265 |
## Model Card Contact
|
| 266 |
|
| 267 |
+
Matt Stammers
|
| 268 |
|
| 269 |
|