Upload 8 files
Browse files- README.md +99 -3
- added_tokens.json +3 -0
- config.json +41 -0
- model.safetensors +3 -0
- special_tokens_map.json +15 -0
- spm.model +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +58 -0
README.md
CHANGED
@@ -1,3 +1,99 @@
|
|
1 |
-
---
|
2 |
-
license: apache-2.0
|
3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
datasets:
|
4 |
+
- agentlans/twitter-sentiment-meta-analysis
|
5 |
+
language:
|
6 |
+
- en
|
7 |
+
base_model:
|
8 |
+
- microsoft/deberta-v3-xsmall
|
9 |
+
- microsoft/deberta-v3-base
|
10 |
+
pipeline_tag: text-classification
|
11 |
+
---
|
12 |
+
# DeBERTa-v3 Twitter Sentiment Models
|
13 |
+
|
14 |
+
This page contains one of two DeBERTa-v3 models (xsmall and base) fine-tuned for Twitter sentiment regression.
|
15 |
+
|
16 |
+
## Model Details
|
17 |
+
|
18 |
+
- **Model Architecture**: DeBERTa-v3
|
19 |
+
- **Variants**:
|
20 |
+
- xsmall (22M parameters)
|
21 |
+
- base (86M parameters)
|
22 |
+
- **Task**: Sentiment regression
|
23 |
+
- **Language**: English
|
24 |
+
- **License**: Apache 2.0
|
25 |
+
|
26 |
+
## Intended Use
|
27 |
+
|
28 |
+
These models are designed for fine-grained sentiment analysis of English tweets. They output a **continuous sentiment score** rather than discrete categories.
|
29 |
+
- negative score means negative sentiment
|
30 |
+
- zero score means neutral sentiment
|
31 |
+
- positive score means positive sentiment
|
32 |
+
- the absolute value of the score represents how strong that sentiment is
|
33 |
+
|
34 |
+
## Training Data
|
35 |
+
|
36 |
+
The models were fine-tuned on a dataset of English tweets collected between September 2009 and January 2010. The sentiment scores were derived from a meta-analysis of 10 different sentiment classifiers using principal component analysis. Find the dataset at [agentlans/twitter-sentiment-meta-analysis](https://huggingface.co/datasets/agentlans/twitter-sentiment-meta-analysis).
|
37 |
+
|
38 |
+
## How to use
|
39 |
+
|
40 |
+
```python
|
41 |
+
from transformers import AutoTokenizer, AutoModelForSequenceClassification
|
42 |
+
import torch
|
43 |
+
|
44 |
+
model_name="agentlans/deberta-v3-base-tweet-sentiment"
|
45 |
+
|
46 |
+
# Put model on GPU or else CPU
|
47 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
48 |
+
model = AutoModelForSequenceClassification.from_pretrained(model_name)
|
49 |
+
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
|
50 |
+
model = model.to(device)
|
51 |
+
|
52 |
+
def sentiment(text):
|
53 |
+
"""Processes the text using the model and returns its logits.
|
54 |
+
In this case, it's interpreted as the sentiment score for that text."""
|
55 |
+
inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True).to(device)
|
56 |
+
with torch.no_grad():
|
57 |
+
logits = model(**inputs).logits.squeeze().cpu()
|
58 |
+
return logits.tolist()
|
59 |
+
|
60 |
+
# Example usage
|
61 |
+
text = [x.strip() for x in """
|
62 |
+
I absolutely despise this product and regret ever purchasing it.
|
63 |
+
The service at that restaurant was terrible and ruined our entire evening.
|
64 |
+
I'm feeling a bit under the weather today, but it's not too bad.
|
65 |
+
The weather is quite average today, neither good nor bad.
|
66 |
+
The movie was okay, I didn't love it but I didn't hate it either.
|
67 |
+
I'm looking forward to the weekend, it should be nice to relax.
|
68 |
+
This new coffee shop has a really pleasant atmosphere and friendly staff.
|
69 |
+
I'm thrilled with my new job and the opportunities it presents!
|
70 |
+
The concert last night was absolutely incredible, easily the best I've ever seen.
|
71 |
+
I'm overjoyed and grateful for all the love and support from my friends and family.
|
72 |
+
""".strip().split("\n")]
|
73 |
+
|
74 |
+
for x, s in zip(text, sentiment(text)):
|
75 |
+
print(f"Text: {x}\nSentiment: {round(s, 2)}\n")
|
76 |
+
```
|
77 |
+
|
78 |
+
Output:
|
79 |
+
```text
|
80 |
+
|
81 |
+
```
|
82 |
+
|
83 |
+
## Performance
|
84 |
+
|
85 |
+
Evaluation set RMSE:
|
86 |
+
- xsmall: 0.2560
|
87 |
+
- base: 0.1938
|
88 |
+
|
89 |
+
## Limitations
|
90 |
+
|
91 |
+
- English language only
|
92 |
+
- Trained specifically on tweets, may or may not generalize well to other text types
|
93 |
+
- Lack of broader context beyond individual tweets
|
94 |
+
- May struggle with detecting sarcasm or nuanced sentiment
|
95 |
+
|
96 |
+
## Ethical Considerations
|
97 |
+
|
98 |
+
- Potential biases in the training data related to the time period and Twitter user demographics
|
99 |
+
- Risk of misuse for large-scale sentiment monitoring without consent
|
added_tokens.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"[MASK]": 128000
|
3 |
+
}
|
config.json
ADDED
@@ -0,0 +1,41 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "microsoft/deberta-v3-base",
|
3 |
+
"architectures": [
|
4 |
+
"DebertaV2ForSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"hidden_act": "gelu",
|
8 |
+
"hidden_dropout_prob": 0.1,
|
9 |
+
"hidden_size": 768,
|
10 |
+
"id2label": {
|
11 |
+
"0": "LABEL_0"
|
12 |
+
},
|
13 |
+
"initializer_range": 0.02,
|
14 |
+
"intermediate_size": 3072,
|
15 |
+
"label2id": {
|
16 |
+
"LABEL_0": 0
|
17 |
+
},
|
18 |
+
"layer_norm_eps": 1e-07,
|
19 |
+
"max_position_embeddings": 512,
|
20 |
+
"max_relative_positions": -1,
|
21 |
+
"model_type": "deberta-v2",
|
22 |
+
"norm_rel_ebd": "layer_norm",
|
23 |
+
"num_attention_heads": 12,
|
24 |
+
"num_hidden_layers": 12,
|
25 |
+
"pad_token_id": 0,
|
26 |
+
"pooler_dropout": 0,
|
27 |
+
"pooler_hidden_act": "gelu",
|
28 |
+
"pooler_hidden_size": 768,
|
29 |
+
"pos_att_type": [
|
30 |
+
"p2c",
|
31 |
+
"c2p"
|
32 |
+
],
|
33 |
+
"position_biased_input": false,
|
34 |
+
"position_buckets": 256,
|
35 |
+
"relative_attention": true,
|
36 |
+
"share_att_key": true,
|
37 |
+
"torch_dtype": "float32",
|
38 |
+
"transformers_version": "4.45.1",
|
39 |
+
"type_vocab_size": 0,
|
40 |
+
"vocab_size": 128100
|
41 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5b7123eaa1dd5d6fe67bd63d00506646ff0a621436c9984fde716693f6951353
|
3 |
+
size 737716196
|
special_tokens_map.json
ADDED
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"bos_token": "[CLS]",
|
3 |
+
"cls_token": "[CLS]",
|
4 |
+
"eos_token": "[SEP]",
|
5 |
+
"mask_token": "[MASK]",
|
6 |
+
"pad_token": "[PAD]",
|
7 |
+
"sep_token": "[SEP]",
|
8 |
+
"unk_token": {
|
9 |
+
"content": "[UNK]",
|
10 |
+
"lstrip": false,
|
11 |
+
"normalized": true,
|
12 |
+
"rstrip": false,
|
13 |
+
"single_word": false
|
14 |
+
}
|
15 |
+
}
|
spm.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
|
3 |
+
size 2464616
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"added_tokens_decoder": {
|
3 |
+
"0": {
|
4 |
+
"content": "[PAD]",
|
5 |
+
"lstrip": false,
|
6 |
+
"normalized": false,
|
7 |
+
"rstrip": false,
|
8 |
+
"single_word": false,
|
9 |
+
"special": true
|
10 |
+
},
|
11 |
+
"1": {
|
12 |
+
"content": "[CLS]",
|
13 |
+
"lstrip": false,
|
14 |
+
"normalized": false,
|
15 |
+
"rstrip": false,
|
16 |
+
"single_word": false,
|
17 |
+
"special": true
|
18 |
+
},
|
19 |
+
"2": {
|
20 |
+
"content": "[SEP]",
|
21 |
+
"lstrip": false,
|
22 |
+
"normalized": false,
|
23 |
+
"rstrip": false,
|
24 |
+
"single_word": false,
|
25 |
+
"special": true
|
26 |
+
},
|
27 |
+
"3": {
|
28 |
+
"content": "[UNK]",
|
29 |
+
"lstrip": false,
|
30 |
+
"normalized": true,
|
31 |
+
"rstrip": false,
|
32 |
+
"single_word": false,
|
33 |
+
"special": true
|
34 |
+
},
|
35 |
+
"128000": {
|
36 |
+
"content": "[MASK]",
|
37 |
+
"lstrip": false,
|
38 |
+
"normalized": false,
|
39 |
+
"rstrip": false,
|
40 |
+
"single_word": false,
|
41 |
+
"special": true
|
42 |
+
}
|
43 |
+
},
|
44 |
+
"bos_token": "[CLS]",
|
45 |
+
"clean_up_tokenization_spaces": false,
|
46 |
+
"cls_token": "[CLS]",
|
47 |
+
"do_lower_case": false,
|
48 |
+
"eos_token": "[SEP]",
|
49 |
+
"mask_token": "[MASK]",
|
50 |
+
"model_max_length": 1000000000000000019884624838656,
|
51 |
+
"pad_token": "[PAD]",
|
52 |
+
"sep_token": "[SEP]",
|
53 |
+
"sp_model_kwargs": {},
|
54 |
+
"split_by_punct": false,
|
55 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
56 |
+
"unk_token": "[UNK]",
|
57 |
+
"vocab_type": "spm"
|
58 |
+
}
|