ybelkada commited on
Commit
da3dc27
1 Parent(s): e11f578

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +24 -12
README.md CHANGED
@@ -64,10 +64,22 @@ language:
64
  widget:
65
  - text: "Translate to German: My name is Arthur"
66
  example_title: "Translation"
67
- - text: "Please answer to the following question. Who is going to be the nextBallon d'or’?"
68
  example_title: "Question Answering"
69
- - text: "Answer the following question by reasoning step-by-step. The cafeteria had 23 apples. If they used 20 for lunch and bought 6 more, how many apples do they have?"
70
- example_title: "Logic puzzles"
 
 
 
 
 
 
 
 
 
 
 
 
71
 
72
  tags:
73
  - text2text-generation
@@ -142,8 +154,8 @@ Find below some example scripts on how to use the model in `transformers`:
142
 
143
  from transformers import T5Tokenizer, T5ForConditionalGeneration
144
 
145
- tokenizer = T5Tokenizer.from_pretrained("google/flan-t5-large")
146
- model = T5ForConditionalGeneration.from_pretrained("google/flan-t5-large")
147
 
148
  input_text = "translate English to German: How old are you?"
149
  input_ids = tokenizer.encode(input_text, return_tensors="pt").input_ids
@@ -163,8 +175,8 @@ print(tokenizer.decode(outputs[0]))
163
  # pip install accelerate
164
  from transformers import T5Tokenizer, T5ForConditionalGeneration
165
 
166
- tokenizer = T5Tokenizer.from_pretrained("google/flan-t5-large")
167
- model = T5ForConditionalGeneration.from_pretrained("google/flan-t5-large", device_map="auto")
168
 
169
  input_text = "translate English to German: How old are you?"
170
  input_ids = tokenizer.encode(input_text, return_tensors="pt").input_ids.to("cuda")
@@ -187,8 +199,8 @@ print(tokenizer.decode(outputs[0]))
187
  import torch
188
  from transformers import T5Tokenizer, T5ForConditionalGeneration
189
 
190
- tokenizer = T5Tokenizer.from_pretrained("google/flan-t5-large")
191
- model = T5ForConditionalGeneration.from_pretrained("google/flan-t5-large", device_map="auto", torch_dtype=torch.float16)
192
 
193
  input_text = "translate English to German: How old are you?"
194
  input_ids = tokenizer.encode(input_text, return_tensors="pt").input_ids.to("cuda")
@@ -208,8 +220,8 @@ print(tokenizer.decode(outputs[0]))
208
  # pip install bitsandbytes accelerate
209
  from transformers import T5Tokenizer, T5ForConditionalGeneration
210
 
211
- tokenizer = T5Tokenizer.from_pretrained("google/flan-t5-large")
212
- model = T5ForConditionalGeneration.from_pretrained("google/flan-t5-large", device_map="auto", load_in_8bit=True)
213
 
214
  input_text = "translate English to German: How old are you?"
215
  input_ids = tokenizer.encode(input_text, return_tensors="pt").input_ids.to("cuda")
@@ -280,7 +292,7 @@ For full details, please check the [research paper](https://arxiv.org/pdf/2210.1
280
 
281
  ## Results
282
 
283
- For full results for FLAN-T5-Large, see the [research paper](https://arxiv.org/pdf/2210.11416.pdf), Table 3.
284
 
285
  # Environmental Impact
286
 
 
64
  widget:
65
  - text: "Translate to German: My name is Arthur"
66
  example_title: "Translation"
67
+ - text: "Please answer to the following question. Who is going to be the next Ballon d'or?"
68
  example_title: "Question Answering"
69
+ - text: "Q: Can Geoffrey Hinton have a conversation with George Washington? Give the rationale before answering."
70
+ example_title: "Logical reasoning"
71
+ - text: "Please answer the following question. What is the boiling point of Nitrogen?"
72
+ example_title: "Scientific knowledge"
73
+ - text: "Answer the following yes/no question. Can you write a whole Haiku in a single tweet?"
74
+ example_title: "Yes/no question"
75
+ - text: "Answer the following yes/no question by reasoning step-by-step. Can you write a whole Haiku in a single tweet?"
76
+ example_title: "Reasoning task"
77
+ - text: "Q: ( False or not False or False ) is? A: Let's think step by step"
78
+ example_title: "Boolean Expressions"
79
+ - text: "The square root of x is the cube root of y. What is y to the power of 2, if x = 4?"
80
+ example_title: "Math reasoning"
81
+ - text: "Premise: At my age you will probably have learnt one lesson. Hypothesis: It's not certain how many lessons you'll learn by your thirties. Does the premise entail the hypothesis?"
82
+ example_title: "Premise and hypothesis"
83
 
84
  tags:
85
  - text2text-generation
 
154
 
155
  from transformers import T5Tokenizer, T5ForConditionalGeneration
156
 
157
+ tokenizer = T5Tokenizer.from_pretrained("google/flan-t5-xxl")
158
+ model = T5ForConditionalGeneration.from_pretrained("google/flan-t5-xxl")
159
 
160
  input_text = "translate English to German: How old are you?"
161
  input_ids = tokenizer.encode(input_text, return_tensors="pt").input_ids
 
175
  # pip install accelerate
176
  from transformers import T5Tokenizer, T5ForConditionalGeneration
177
 
178
+ tokenizer = T5Tokenizer.from_pretrained("google/flan-t5-xxl")
179
+ model = T5ForConditionalGeneration.from_pretrained("google/flan-t5-xxl", device_map="auto")
180
 
181
  input_text = "translate English to German: How old are you?"
182
  input_ids = tokenizer.encode(input_text, return_tensors="pt").input_ids.to("cuda")
 
199
  import torch
200
  from transformers import T5Tokenizer, T5ForConditionalGeneration
201
 
202
+ tokenizer = T5Tokenizer.from_pretrained("google/flan-t5-xxl")
203
+ model = T5ForConditionalGeneration.from_pretrained("google/flan-t5-xxl", device_map="auto", torch_dtype=torch.float16)
204
 
205
  input_text = "translate English to German: How old are you?"
206
  input_ids = tokenizer.encode(input_text, return_tensors="pt").input_ids.to("cuda")
 
220
  # pip install bitsandbytes accelerate
221
  from transformers import T5Tokenizer, T5ForConditionalGeneration
222
 
223
+ tokenizer = T5Tokenizer.from_pretrained("google/flan-t5-xxl")
224
+ model = T5ForConditionalGeneration.from_pretrained("google/flan-t5-xxl", device_map="auto", load_in_8bit=True)
225
 
226
  input_text = "translate English to German: How old are you?"
227
  input_ids = tokenizer.encode(input_text, return_tensors="pt").input_ids.to("cuda")
 
292
 
293
  ## Results
294
 
295
+ For full results for FLAN-T5-XXL, see the [research paper](https://arxiv.org/pdf/2210.11416.pdf), Table 3.
296
 
297
  # Environmental Impact
298