msgfrom96/xlm_emo_multi

README.md

@@ -1,7 +1,7 @@
 ---
 library_name: transformers
 license: mit
-base_model: xlm-roberta-base
+base_model: xlm-roberta-large
 tags:
 - generated_from_trainer
 model-index:
@@ -14,11 +14,11 @@ should probably proofread and complete it, then remove this comment. -->
 
 # emotion_model
 
-This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the None dataset.
+This model is a fine-tuned version of [xlm-roberta-large](https://huggingface.co/xlm-roberta-large) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3094
-- Micro F1: 0.6854
-- Macro F1: 0.4971
+- Loss: 0.2936
+- Micro F1: 0.7177
+- Macro F1: 0.5792
 
 ## Model description
 
@@ -51,16 +51,16 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss | Micro F1 | Macro F1 |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|:--------:|
-| 0.5206        | 1.0   | 143  | 0.4242          | 0.3130   | 0.1419   |
-| 0.3877        | 2.0   | 286  | 0.3661          | 0.6053   | 0.3646   |
-| 0.3538        | 3.0   | 429  | 0.3528          | 0.6106   | 0.3874   |
-| 0.3306        | 4.0   | 572  | 0.3341          | 0.6563   | 0.4442   |
-| 0.3055        | 5.0   | 715  | 0.3254          | 0.6761   | 0.4663   |
-| 0.2865        | 6.0   | 858  | 0.3222          | 0.6776   | 0.4853   |
-| 0.2811        | 7.0   | 1001 | 0.3169          | 0.6804   | 0.4766   |
-| 0.2718        | 8.0   | 1144 | 0.3154          | 0.6808   | 0.4874   |
-| 0.2673        | 9.0   | 1287 | 0.3175          | 0.6802   | 0.4922   |
-| 0.2646        | 10.0  | 1430 | 0.3166          | 0.6827   | 0.4955   |
+| 0.4725        | 1.0   | 143  | 0.3610          | 0.5857   | 0.3394   |
+| 0.3219        | 2.0   | 286  | 0.3058          | 0.6957   | 0.5022   |
+| 0.2892        | 3.0   | 429  | 0.3067          | 0.6814   | 0.4921   |
+| 0.263         | 4.0   | 572  | 0.2915          | 0.7037   | 0.5397   |
+| 0.2399        | 5.0   | 715  | 0.2911          | 0.7108   | 0.5625   |
+| 0.2255        | 6.0   | 858  | 0.2922          | 0.7115   | 0.5833   |
+| 0.2147        | 7.0   | 1001 | 0.2877          | 0.7115   | 0.5792   |
+| 0.1989        | 8.0   | 1144 | 0.2922          | 0.7176   | 0.5882   |
+| 0.1908        | 9.0   | 1287 | 0.2956          | 0.7162   | 0.5816   |
+| 0.1851        | 10.0  | 1430 | 0.2970          | 0.7193   | 0.5984   |
 
 
 ### Framework versions

config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "xlm-roberta-base",
+  "_name_or_path": "xlm-roberta-large",
   "architectures": [
     "XLMRobertaForSequenceClassification"
   ],
@@ -9,7 +9,7 @@
   "eos_token_id": 2,
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.1,
-  "hidden_size": 768,
+  "hidden_size": 1024,
   "id2label": {
     "0": "anger",
     "1": "anticipation",
@@ -24,7 +24,7 @@
     "10": "pessimism"
   },
   "initializer_range": 0.02,
-  "intermediate_size": 3072,
+  "intermediate_size": 4096,
   "label2id": {
     "anger": 0,
     "anticipation": 1,
@@ -41,8 +41,8 @@
   "layer_norm_eps": 1e-05,
   "max_position_embeddings": 514,
   "model_type": "xlm-roberta",
-  "num_attention_heads": 12,
-  "num_hidden_layers": 12,
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
   "output_past": true,
   "pad_token_id": 1,
   "position_embedding_type": "absolute",

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:23b7b59a069ea154dc7742fed8314f9f7829a49028f5ed20680ab93a93609b41
-size 1112232692
+oid sha256:91a64034f69d6333a6584e016aa189cf01372ef99c0d34c1ea8d05edb0cd8b4a
+size 2239655572

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6bdb7b00eeb43cb37cf9061bad72e12e52eb40b67798eb96e4b3f4ca90ccea54
+oid sha256:848316f869eb3b1363d9052b1f9bfe54b5c795f4a47506ae95e4a324cb5266dc
 size 5304

xlm_multi_emo/config.json

@@ -1,5 +1,5 @@
 {
-  "_name_or_path": "xlm-roberta-base",
+  "_name_or_path": "xlm-roberta-large",
   "architectures": [
     "XLMRobertaForSequenceClassification"
   ],
@@ -9,7 +9,7 @@
   "eos_token_id": 2,
   "hidden_act": "gelu",
   "hidden_dropout_prob": 0.1,
-  "hidden_size": 768,
+  "hidden_size": 1024,
   "id2label": {
     "0": "anger",
     "1": "anticipation",
@@ -24,7 +24,7 @@
     "10": "pessimism"
   },
   "initializer_range": 0.02,
-  "intermediate_size": 3072,
+  "intermediate_size": 4096,
   "label2id": {
     "anger": 0,
     "anticipation": 1,
@@ -41,8 +41,8 @@
   "layer_norm_eps": 1e-05,
   "max_position_embeddings": 514,
   "model_type": "xlm-roberta",
-  "num_attention_heads": 12,
-  "num_hidden_layers": 12,
+  "num_attention_heads": 16,
+  "num_hidden_layers": 24,
   "output_past": true,
   "pad_token_id": 1,
   "position_embedding_type": "absolute",

xlm_multi_emo/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:23b7b59a069ea154dc7742fed8314f9f7829a49028f5ed20680ab93a93609b41
-size 1112232692
+oid sha256:91a64034f69d6333a6584e016aa189cf01372ef99c0d34c1ea8d05edb0cd8b4a
+size 2239655572

xlm_multi_emo/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6bdb7b00eeb43cb37cf9061bad72e12e52eb40b67798eb96e4b3f4ca90ccea54
+oid sha256:848316f869eb3b1363d9052b1f9bfe54b5c795f4a47506ae95e4a324cb5266dc
 size 5304