anyone here worked with dream shaper text to image model ? i tried to get some result from it but i

M

MohamadOP•3/3/25, 6:55 AM

to my experience they are using a wrong sampler on the server side , so cloudfalre doens't allow us to cahnge the sampler , and so this is the WRONG result we are getting from this model

V

Victor•3/4/25, 12:04 AM

The model page

llama-guard-3-8b

llama-guard-3-8b

llama-guard-3-8b

llama-guard-3-8b input schema is just a copy paste of every other model page but it doesn't actually match what is live. For example:

```
system
```
```
system
```
and
```
tool
```
tool
```
tool
```
tool
are not valid message roles

T

tom•3/4/25, 3:14 PM

is there any way to run

pyannote

pyannote

pyannote

pyannote on cloudflare?

云

云•3/5/25, 10:00 PM

云

云•3/5/25, 10:00 PM

bruh

Ttom is there any way to run `pyannote` on cloudflare?

T

tornado•3/5/25, 10:48 PM

It's a heavy model, I would be surprised if they add it

T

tornado•3/5/25, 10:50 PM

Can we get a VAD model to workers AI? The whisper model have vad so I imagine having a just VAD model can be done easily and can be very useful for voice calls.

N

nclevenger•3/6/25, 11:39 AM

Would love this new reasoning model released yesterday on Workers AI ... it looks like it substantially outperforms the DeepSeek R1 Qwen 32b distillation: https://huggingface.co/Qwen/QwQ-32B

Qwen/QwQ-32B · Hugging Face

Nnclevenger Would love this new reasoning model released yesterday on Workers AI ... it look...

F

fatwang2•3/7/25, 2:16 AM

why I can't find it on workers ai?

Ffatwang2 why I can't find it on workers ai?

I

Isaac McFadyen•3/7/25, 3:40 AM

It's not on Workers AI yet.

I

Isaac McFadyen•3/7/25, 3:40 AM

They are requesting it be added, not saying it's already on Workers AI.

MMohamad anyone here worked with dream shaper text to image model ? i tried to get some ...

R

Raylight•3/7/25, 9:50 AM

The only way I been able to get decent results is by setting

guidance

guidance

guidance

guidance to something like 1, possibly 2. With higher values, problems with saturation and contrast quickly become apparent. The img2img and inpainting models had similar issues last time I checked. (See cloudflare-ai for example.) And SDXL-lightning has a somewhat similar issue. (Cartoonish speckled result unless you set

num_steps

num_steps

num_steps

num_steps to e.g. 2 and

guidance

guidance

guidance

guidance to a low value.)

L

lfdepombo•3/7/25, 4:51 PM

Hi, I am consistently having requests through the

ai/run

ai/run

ai/run

ai/run endpoint truncated when

@cf/deepseek-ai/deepseek-r1-distill-qwen-32b

@cf/deepseek-ai/deepseek-r1-distill-qwen-32b

@cf/deepseek-ai/deepseek-r1-distill-qwen-32b

@cf/deepseek-ai/deepseek-r1-distill-qwen-32b is clearly midsentence. The request is always 200 too. Is this a known issue of some kind? This has been happening for the past 2 days now.

L

lfdepombo•3/7/25, 4:52 PM

would streaming or using the OpenAI compatible API help at all?

L

lfdepombo•3/7/25, 5:01 PM

ah! default

max_tokens

max_tokens

max_tokens

max_tokens for the result is only 256. bumping it up fixed it. there is no indication about it in the response though

Llfdepombo ah! default `max_tokens` for the result is only 256. bumping it up fixed it. the...

J

James•3/7/25, 5:07 PM

Yeah this was an undocumented breaking change. Previously the defaults weren't enforced but they have been for a short while now. We're trying to get a changelog entry or something about this so folks are aware of the change.

A

ajgeiss0702•3/7/25, 11:52 PM

it looks like

@cf/deepseek-ai/deepseek-r1-distill-qwen-32b

@cf/deepseek-ai/deepseek-r1-distill-qwen-32b

@cf/deepseek-ai/deepseek-r1-distill-qwen-32b

@cf/deepseek-ai/deepseek-r1-distill-qwen-32b is missing the starting

<think>

<think>

<think>

<think> tag in responses

G

Gyurmatag•3/8/25, 11:42 AM

Hi guys!
Should't we add a new channel inside AI+ML for the Agents SDK?

K

kian•3/8/25, 4:55 PM

#agents

K

kian•3/8/25, 4:55 PM

There is already one

R

Rubi•3/9/25, 10:24 AM

why cloudflare deepseek qwen is not compatible like Deepseek official cloud?

Aajgeiss0702 it looks like `@cf/deepseek-ai/deepseek-r1-distill-qwen-32b` is missing the star...

R

Rubi•3/9/25, 1:29 PM

yeah why this happen

L

Love•3/10/25, 10:23 PM

Hey, I'm completely new to Cloudflare use but I'm setting up a worker for parsing JSON to my discord bot but im having difficulties setting up the worker and "connecting it" to my docker runtime

R

Rubi•3/11/25, 8:04 AM

you can now use Cloudflare R1 Distill Qwen with browser-use

https://github.com/browser-use/browser-use

GitHub

GitHub - browser-use/browser-use: Make websites accessible for AI a...

Make websites accessible for AI agents. Contribute to browser-use/browser-use development by creating an account on GitHub.

M

Muhammadreza Haghiri•3/11/25, 8:58 AM

I am using FLUX Schnell like this:

curl --location --request POST 'https://api.cloudflare.com/client/v4/accounts/REDACTED/ai/run/@cf/black-forest-labs/flux-1-schnell' \
--header 'Authorization: Bearer REDACTED' \
--header 'Content-Type: application/json' \
--header 'Cookie: __cf_bm=V5sLTJx2Y7d6s6bClk.umX36FohdsTMoXamYf1pW1f8-1741683179-1.0.1.1-rGuk5rHKjPNKMCBtB45ZW51MZeJUyqVGXeW4hr0o0m7F7JFsT315BeJAbXd8LKCOuApweso_4kAYQf4SfM6R60cd.ZtvxRyn6o3qLISOqS4' \
--data-raw '{
    "prompt" : "A serene tropical beach at sunset, a cozy wooden cottage with smoke rising from the chimney, tall swaying palm trees, calm ocean waves reflecting golden hues of sunlight, painted in a dreamy watercolor style with soft, vibrant colors and intricate details.", 
    "steps" : 8,
    "width" :1024, 
    "height" : 1024, 
    "seed" : 93617911245
}'

curl --location --request POST 'https://api.cloudflare.com/client/v4/accounts/REDACTED/ai/run/@cf/black-forest-labs/flux-1-schnell' \
--header 'Authorization: Bearer REDACTED' \
--header 'Content-Type: application/json' \
--header 'Cookie: __cf_bm=V5sLTJx2Y7d6s6bClk.umX36FohdsTMoXamYf1pW1f8-1741683179-1.0.1.1-rGuk5rHKjPNKMCBtB45ZW51MZeJUyqVGXeW4hr0o0m7F7JFsT315BeJAbXd8LKCOuApweso_4kAYQf4SfM6R60cd.ZtvxRyn6o3qLISOqS4' \
--data-raw '{
    "prompt" : "A serene tropical beach at sunset, a cozy wooden cottage with smoke rising from the chimney, tall swaying palm trees, calm ocean waves reflecting golden hues of sunlight, painted in a dreamy watercolor style with soft, vibrant colors and intricate details.", 
    "steps" : 8,
    "width" :1024, 
    "height" : 1024, 
    "seed" : 93617911245
}'

curl --location --request POST 'https://api.cloudflare.com/client/v4/accounts/REDACTED/ai/run/@cf/black-forest-labs/flux-1-schnell' \
--header 'Authorization: Bearer REDACTED' \
--header 'Content-Type: application/json' \
--header 'Cookie: __cf_bm=V5sLTJx2Y7d6s6bClk.umX36FohdsTMoXamYf1pW1f8-1741683179-1.0.1.1-rGuk5rHKjPNKMCBtB45ZW51MZeJUyqVGXeW4hr0o0m7F7JFsT315BeJAbXd8LKCOuApweso_4kAYQf4SfM6R60cd.ZtvxRyn6o3qLISOqS4' \
--data-raw '{
    "prompt" : "A serene tropical beach at sunset, a cozy wooden cottage with smoke rising from the chimney, tall swaying palm trees, calm ocean waves reflecting golden hues of sunlight, painted in a dreamy watercolor style with soft, vibrant colors and intricate details.", 
    "steps" : 8,
    "width" :1024, 
    "height" : 1024, 
    "seed" : 93617911245
}'

curl --location --request POST 'https://api.cloudflare.com/client/v4/accounts/REDACTED/ai/run/@cf/black-forest-labs/flux-1-schnell' \
--header 'Authorization: Bearer REDACTED' \
--header 'Content-Type: application/json' \
--header 'Cookie: __cf_bm=V5sLTJx2Y7d6s6bClk.umX36FohdsTMoXamYf1pW1f8-1741683179-1.0.1.1-rGuk5rHKjPNKMCBtB45ZW51MZeJUyqVGXeW4hr0o0m7F7JFsT315BeJAbXd8LKCOuApweso_4kAYQf4SfM6R60cd.ZtvxRyn6o3qLISOqS4' \
--data-raw '{
    "prompt" : "A serene tropical beach at sunset, a cozy wooden cottage with smoke rising from the chimney, tall swaying palm trees, calm ocean waves reflecting golden hues of sunlight, painted in a dreamy watercolor style with soft, vibrant colors and intricate details.", 
    "steps" : 8,
    "width" :1024, 
    "height" : 1024, 
    "seed" : 93617911245
}'

And I get this error:

{
    "errors": [
        {
            "message": "AiError: AiError: Invalid input",
            "code": 5006
        }
    ],
    "success": false,
    "result": {},
    "messages": []
}

{
    "errors": [
        {
            "message": "AiError: AiError: Invalid input",
            "code": 5006
        }
    ],
    "success": false,
    "result": {},
    "messages": []
}

{
    "errors": [
        {
            "message": "AiError: AiError: Invalid input",
            "code": 5006
        }
    ],
    "success": false,
    "result": {},
    "messages": []
}

{
    "errors": [
        {
            "message": "AiError: AiError: Invalid input",
            "code": 5006
        }
    ],
    "success": false,
    "result": {},
    "messages": []
}

What causes this? What modification should I make to my code?

M

Muhammadreza Haghiri•3/11/25, 9:49 AM

I looked at the docs and there is no error code for 5006.

M

Muhammadreza Haghiri•3/11/25, 2:50 PM

It works once and doesn't for a 100 times.

C

Chaika•3/11/25, 3:13 PM

Thanks, repro'd and raising

CChaika Thanks, repro'd and raising

M

Muhammadreza Haghiri•3/11/25, 3:53 PM

Was it for my problem?

MMuhammadreza Haghiri Was it for my problem?

C

Chaika•3/11/25, 3:59 PM

Yes. CF acknowledged the report, hopefully fix soon, seems to have been erroring for a bit..

CChaika Yes. CF acknowledged the report, hopefully fix soon, seems to have been erroring...

M

Muhammadreza Haghiri•3/11/25, 4:12 PM

Thanks

D

dan•3/11/25, 6:08 PM

are there any STT models on workers ai that support live streaming?

Ddan are there any STT models on workers ai that support live streaming?

I

Isaac McFadyen•3/11/25, 6:16 PM

No, and a lot of the new models being developed don't stream at all any more, actually. For example, Whisper/Nvidia Parakeet/Nvidia Canary don't support streaming, and instead recommend applications implement their own streaming by sending a request every <n> seconds while continuing recording (e.g. using multiple threads).

MMuhammadreza Haghiri I am using FLUX Schnell like this: ``` curl --location --request POST 'https:/...

A

ale•3/11/25, 6:26 PM

hi @Muhammadreza Haghiri, could you please try this request once again? The team has deployed a fix

D

DennisMotta•3/12/25, 12:51 AM

why language auto-detection is not supported on whisper-large-v3-turbo?

D

DennisMotta•3/12/25, 12:51 AM

by default it transcribes to English, I would like to keep the original language, similar to what @cf/openai/whisper does

M

Max•3/12/25, 9:12 AM

Hi. is this the correct channel to ask about workers-mcp? if so: what is the ETA on

Remote MCP support: Connect to MCP servers directly without requiring a local proxy

Remote MCP support: Connect to MCP servers directly without requiring a local proxy

Remote MCP support: Connect to MCP servers directly without requiring a local proxy

Remote MCP support: Connect to MCP servers directly without requiring a local proxy @geleen

D

dmaretskyi•3/12/25, 2:18 PM

Does

deepseek-r1-distill-qwen-32b

deepseek-r1-distill-qwen-32b

deepseek-r1-distill-qwen-32b

deepseek-r1-distill-qwen-32b support tools calls in stream mode? Is there a code example?

Aale hi @Muhammadreza Haghiri, could you please try this request once again? The team...

M

Muhammadreza Haghiri•3/12/25, 7:33 PM

Hi. It was okay today.

LLiam Every single day :/ The fail rate at this point is just 99%

C

certified dirtiboi•3/12/25, 10:24 PM

Also seeing this error. I've seen other posts saying it's a network issue at the workers-ai level, but I have yet to see a resolution or RCA

Is the solution to just add a fallback retry?

errors: [
{
message: 'AiError: AiError: Capacity temporarily exceeded, please try again.',
code: 3040
}
]

Aajgeiss0702 it looks like `@cf/deepseek-ai/deepseek-r1-distill-qwen-32b` is missing the star...

A

ajgeiss0702•3/13/25, 12:35 AM

bump

Ddmaretskyi Does `deepseek-r1-distill-qwen-32b` support tools calls in stream mode? Is there...

D

dmaretskyi•3/13/25, 5:27 PM

Here's the snipped I'm running;

const response = await env.AI.run('@cf/deepseek-ai/deepseek-r1-distill-qwen-32b', {
    messages: [
      {
        role: 'user',
        content: 'Make a cup of earl grey tea. Use the tool provided.',
      },
    ],
    tools: [
      {
        name: 'make_tea',
        description: 'Make a cup of tea',
        parameters: {
          type: 'object',
          properties: {
            tea_type: {
              type: 'string',
              description: 'The type of tea to make',
              enum: ['earl grey', 'chamomile', 'peppermint'],
            },
          },
        },
      },
    ],
  } satisfies BaseAiTextGeneration['inputs']);

  log.info('generateCloudflare', { response });

const response = await env.AI.run('@cf/deepseek-ai/deepseek-r1-distill-qwen-32b', {
    messages: [
      {
        role: 'user',
        content: 'Make a cup of earl grey tea. Use the tool provided.',
      },
    ],
    tools: [
      {
        name: 'make_tea',
        description: 'Make a cup of tea',
        parameters: {
          type: 'object',
          properties: {
            tea_type: {
              type: 'string',
              description: 'The type of tea to make',
              enum: ['earl grey', 'chamomile', 'peppermint'],
            },
          },
        },
      },
    ],
  } satisfies BaseAiTextGeneration['inputs']);

  log.info('generateCloudflare', { response });

const response = await env.AI.run('@cf/deepseek-ai/deepseek-r1-distill-qwen-32b', {
    messages: [
      {
        role: 'user',
        content: 'Make a cup of earl grey tea. Use the tool provided.',
      },
    ],
    tools: [
      {
        name: 'make_tea',
        description: 'Make a cup of tea',
        parameters: {
          type: 'object',
          properties: {
            tea_type: {
              type: 'string',
              description: 'The type of tea to make',
              enum: ['earl grey', 'chamomile', 'peppermint'],
            },
          },
        },
      },
    ],
  } satisfies BaseAiTextGeneration['inputs']);

  log.info('generateCloudflare', { response });

const response = await env.AI.run('@cf/deepseek-ai/deepseek-r1-distill-qwen-32b', {
    messages: [
      {
        role: 'user',
        content: 'Make a cup of earl grey tea. Use the tool provided.',
      },
    ],
    tools: [
      {
        name: 'make_tea',
        description: 'Make a cup of tea',
        parameters: {
          type: 'object',
          properties: {
            tea_type: {
              type: 'string',
              description: 'The type of tea to make',
              enum: ['earl grey', 'chamomile', 'peppermint'],
            },
          },
        },
      },
    ],
  } satisfies BaseAiTextGeneration['inputs']);

  log.info('generateCloudflare', { response });

 {
  response: {
    response: "Okay, so I need to make a cup of Earl Grey tea using the tools provided. Let me think about what I know about making tea and what tools I might have access to. First, I know that Earl Grey is a black tea with a bergamot flavor, so it's usually served hot and might be drunk with milk or just on its own.\n" +
      '\n' +
      "I should start by figuring out what tools I need. The user mentioned to use the tool provided, but didn't specify what it is. So I'll assume I have a teapot, a kettle for boiling water, a teaspoon for measuring tea, a cup, maybe a strainer, and perhaps a milk pitcher if I want to add milk. I should also consider whether the tools are electric or manual, but I'll go with basic assumptions.\n" +
      '\n' +
      "First step: Boil water. I'll need to fill the kettle with fresh water and bring it to a boil. The temperature for black tea is usually around 90-100°C, but I think boiling water is fine. Wait, no, sometimes it's better to let the water cool a bit after boiling to prevent over-extracting the tea. But Earl Grey is a black tea, so maybe it's okay to use boiling water.\n" +
      '\n',
    usage: { prompt_tokens: 20, completion_tokens: 256, total_tokens: 276 }
  }
}

 {
  response: {
    response: "Okay, so I need to make a cup of Earl Grey tea using the tools provided. Let me think about what I know about making tea and what tools I might have access to. First, I know that Earl Grey is a black tea with a bergamot flavor, so it's usually served hot and might be drunk with milk or just on its own.\n" +
      '\n' +
      "I should start by figuring out what tools I need. The user mentioned to use the tool provided, but didn't specify what it is. So I'll assume I have a teapot, a kettle for boiling water, a teaspoon for measuring tea, a cup, maybe a strainer, and perhaps a milk pitcher if I want to add milk. I should also consider whether the tools are electric or manual, but I'll go with basic assumptions.\n" +
      '\n' +
      "First step: Boil water. I'll need to fill the kettle with fresh water and bring it to a boil. The temperature for black tea is usually around 90-100°C, but I think boiling water is fine. Wait, no, sometimes it's better to let the water cool a bit after boiling to prevent over-extracting the tea. But Earl Grey is a black tea, so maybe it's okay to use boiling water.\n" +
      '\n',
    usage: { prompt_tokens: 20, completion_tokens: 256, total_tokens: 276 }
  }
}

 {
  response: {
    response: "Okay, so I need to make a cup of Earl Grey tea using the tools provided. Let me think about what I know about making tea and what tools I might have access to. First, I know that Earl Grey is a black tea with a bergamot flavor, so it's usually served hot and might be drunk with milk or just on its own.\n" +
      '\n' +
      "I should start by figuring out what tools I need. The user mentioned to use the tool provided, but didn't specify what it is. So I'll assume I have a teapot, a kettle for boiling water, a teaspoon for measuring tea, a cup, maybe a strainer, and perhaps a milk pitcher if I want to add milk. I should also consider whether the tools are electric or manual, but I'll go with basic assumptions.\n" +
      '\n' +
      "First step: Boil water. I'll need to fill the kettle with fresh water and bring it to a boil. The temperature for black tea is usually around 90-100°C, but I think boiling water is fine. Wait, no, sometimes it's better to let the water cool a bit after boiling to prevent over-extracting the tea. But Earl Grey is a black tea, so maybe it's okay to use boiling water.\n" +
      '\n',
    usage: { prompt_tokens: 20, completion_tokens: 256, total_tokens: 276 }
  }
}

 {
  response: {
    response: "Okay, so I need to make a cup of Earl Grey tea using the tools provided. Let me think about what I know about making tea and what tools I might have access to. First, I know that Earl Grey is a black tea with a bergamot flavor, so it's usually served hot and might be drunk with milk or just on its own.\n" +
      '\n' +
      "I should start by figuring out what tools I need. The user mentioned to use the tool provided, but didn't specify what it is. So I'll assume I have a teapot, a kettle for boiling water, a teaspoon for measuring tea, a cup, maybe a strainer, and perhaps a milk pitcher if I want to add milk. I should also consider whether the tools are electric or manual, but I'll go with basic assumptions.\n" +
      '\n' +
      "First step: Boil water. I'll need to fill the kettle with fresh water and bring it to a boil. The temperature for black tea is usually around 90-100°C, but I think boiling water is fine. Wait, no, sometimes it's better to let the water cool a bit after boiling to prevent over-extracting the tea. But Earl Grey is a black tea, so maybe it's okay to use boiling water.\n" +
      '\n',
    usage: { prompt_tokens: 20, completion_tokens: 256, total_tokens: 276 }
  }
}

Ddmaretskyi Here's the snipped I'm running; ```ts const response = await env.AI.run('@cf/dee...

I

Isaac McFadyen•3/13/25, 6:05 PM

Notice how the completion_tokens are 256; the model wasn't actually done, but got cut off because max_tokens defaults to 256 as of a few weeks ago. You'll want to specify a higher maxTokens in your request in order to allow the model to finish it's thought.

J

JoaoCupido•3/13/25, 6:34 PM

Speaking of max_tokens, I can't seem to change max_tokens with the messages property in any chat model, and I have no clue why. I have done this via URL (api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run/@cf/meta/llama-3-8b-instruct).

anyone here worked with dream shaper text to image model ? i tried to get some result from it but i

Similar Threads

Similar Threads

Similar Threads