InvalidRequestError: The model gpt-4-32k

[zimba7768]

I am not sure if this is the correct place to post this; if not, I apologize in advance. I like trying out different AI tools, and currently, I am experimenting with GPT_Author. I have followed all the steps and read as much material as I can find, but I still can't get it to work. I have paid for GPT-4 and for the GPT API. When I run GPT Author, I see the step cost for each section, but in the end, I encounter an error. I am not a programmer, so this is very confusing.

InvalidRequestError Traceback (most recent call last) Cell 7 line 5

3 num_chapters = 15

4 writing_style = "Clear and easily understandable, similar to a young adult novel. Lots of dialogue." ---->

5 novel, title, chapters, chapter_titles = write_fantasy_novel(prompt, num_chapters, writing_style, claude_true)

7 # Replace chapter descriptions with body text in chapter_titles

8 for i, chapter in enumerate(chapters):

Cell 7 line 2

266 chapter_titles = ast.literal_eval(storyline)

269 novel = f"Storyline:\n{storyline}\n\n" -->

271 first_chapter = write_first_chapter(storyline, chapter_titles[0], writing_style.strip(), claude_true) 272 print('first chapter written')

273 novel += f"Chapter 1:\n{first_chapter}\n"

Cell 7 line 1

135 response = openai.ChatCompletion.create(

136 model="gpt-4",

137 messages=[ (...)

140 ]

141 )

143 print_step_costs(response, "gpt-4") -->

145 improved_response = openai.ChatCompletion.create(

...

766 rbody, rcode, resp.data, rheaders, stream_error=stream_error

767 )

768 return resp InvalidRequestError: The model gpt-4-32k does not exist or you do not have access to it. Learn more: https://help.openai.com/en/articles/7102672-how-can-i-access-gpt-4.

[elphin]

I get the same error: InvalidRequestError: The model gpt-4 does not exist or you do not have access to it.

I have a paid openai subscription with an newly generated API key. Didn't work.
Tried it with again a new api key but still get this error.

I have other applications that succesfully use my openai key so that can hardly be the problem.

Anyone?

[StoneyPhenix]

Replace all 'gpt-4-32k' with 'gpt-4' . The 32k is only available to a very small trial group and currently only being given out to the azure clients.

[elphin]

Did just that; replaced every instance but I get the same error...
Any other suggestion?

[zimba7768]

Thank you for replying back to this post. I also did the same suggestion with replacing all the "gpt-4-32K" with "gpt-4". I now got a different error.

KeyError Traceback (most recent call last)
c:\Users\test\gpt-author\gpt_author_v2_1.ipynb Cell 7 line 5
3 num_chapters = 15
4 writing_style = "Clear and easily understandable, similar to a young adult novel. Lots of dialogue."
----> 5 novel, title, chapters, chapter_titles = write_fantasy_novel(prompt, num_chapters, writing_style, claude_true)
7 # Replace chapter descriptions with body text in chapter_titles
8 for i, chapter in enumerate(chapters):

c:\Users\test\gpt-author\gpt_author_v2_1.ipynb Cell 7 line 2
276 for i in range(num_chapters - 1):
277 print(f"Writing chapter {i+2}...") # + 2 because the first chapter was already added
--> 279 chapter = write_chapter(novel, storyline, chapter_titles[i+1])
280 try:
281 if len(str(chapter)) < 100:

c:\Users\test\gpt-author\gpt_author_v2_1.ipynb Cell 7 line 1
170 data = {
171 "model": "claude-2",
172 "prompt": prompt,
173 "max_tokens_to_sample": 5000,
174 }
176 response = requests.post(url, headers=headers, json=data)
--> 178 print_step_costs_anthropic(prompt, response.json()['completion'])
180 return 'Chapter ' + response.json()['completion'].strip().split('```')[0].strip()
181 else:

KeyError: 'completion'

[StoneyPhenix]

That error is happing on anthropic call and I am running without anthropic so I don't know exactly. I would recommended printing out response.json() before trying to extract completion. to see if anthropic is returning an error.

[fuleinist]

I replaced gpt-4-32k with GPT-3.5-Turbo-16k and so far so good

[zimba7768]

I tried to replace gpt-4-32k with GPT-3.5-Turbo-16k and I get this error now. I am sure I am missing something.

InvalidRequestError: The model GPT-3.5-Turbo-16k does not exist

[fuleinist]

I tried to replace gpt-4-32k with GPT-3.5-Turbo-16k and I get this error now. I am sure I am missing something.

InvalidRequestError: The model GPT-3.5-Turbo-16k does not exist

can you paste/screenshot the full error code please?

[zimba7768]

This is what I am getting for an error.

[fuleinist]

This is what I am getting for an error.

can you try "gpt-3.5-turbo-16k" instead of "gpt-3.5-Turbo-16k"?

[simulanics]

You need to use lowercase "gpt-3.5-turbo-16k" - it's case sensitive. Also, In order to access GPT-4, if your PAID (card linked to API billing**) account was created before August 2023, you must spend and have been billed at least $1.00 total in any given month to have been given access to GPT-4 APIs. After August, $0.50. Hope that helps!

[zimba7768]

I did use lowercase and still getting this error.

Step Cost (OpenAI): 0.1167
generated plots
Step Cost (OpenAI): 0.16626000000000002
selected best plot
Step Cost (OpenAI): 0.12078
plot improved
Step Cost (OpenAI): 0.03702
title generated
Generating storyline with chapters and high-level details...
Step Cost (OpenAI): 0.12096
Step Cost (OpenAI): 0.17178
storyline generated
Step Cost (OpenAI): 0.14802
Step Cost (OpenAI): 0.14802
first chapter written
Writing chapter 2...

KeyError Traceback (most recent call last)
c:\Users\zimba\gpt-author\gpt_author_v2_1.ipynb Cell 7 line 5
3 num_chapters = 15
4 writing_style = "Clear and easily understandable, similar to a young adult novel. Lots of dialogue."
----> 5 novel, title, chapters, chapter_titles = write_fantasy_novel(prompt, num_chapters, writing_style, claude_true)
7 # Replace chapter descriptions with body text in chapter_titles
8 for i, chapter in enumerate(chapters):

c:\Users\zimba\gpt-author\gpt_author_v2_1.ipynb Cell 7 line 2
276 for i in range(num_chapters - 1):
277 print(f"Writing chapter {i+2}...") # + 2 because the first chapter was already added
--> 279 chapter = write_chapter(novel, storyline, chapter_titles[i+1])
280 try:
281 if len(str(chapter)) < 100:

c:\Users\zimba\gpt-author\gpt_author_v2_1.ipynb Cell 7 line 1
170 data = {
171 "model": "claude-2",
172 "prompt": prompt,
173 "max_tokens_to_sample": 5000,
174 }
176 response = requests.post(url, headers=headers, json=data)
--> 178 print_step_costs_anthropic(prompt, response.json()['completion'])
180 return 'Chapter ' + response.json()['completion'].strip().split('```')[0].strip()
181 else: