DeepSeek¶

LlamaIndex Llms Integration: DeepSeek¶

This is the DeepSeek integration for LlamaIndex. Visit DeepSeek for information on how to get an API key and which models are supported.

At the time of writing, you can use:

• deepseek-chat
• deepseek-reasoner

Setup¶

If you're opening this Notebook on colab, you will probably need to install LlamaIndex 🦙.

In [ ]:

%pip install llama-index-llms-deepseek

%pip install llama-index-llms-deepseek

In [ ]:

from llama_index.llms.deepseek import DeepSeek

# you can also set DEEPSEEK_API_KEY in your environment variables
llm = DeepSeek(
    model="deepseek-reasoner", api_key="sk-fc30eba000864a32bc8a4bd420100b97"
)

# You might also want to set deepseek as your default llm
# from llama_index.core import Settings
# Settings.llm = llm

from llama_index.llms.deepseek import DeepSeek # you can also set DEEPSEEK_API_KEY in your environment variables llm = DeepSeek( model="deepseek-reasoner", api_key="sk-fc30eba000864a32bc8a4bd420100b97" ) # You might also want to set deepseek as your default llm # from llama_index.core import Settings # Settings.llm = llm

In [ ]:

response = llm.complete("Is 9.9 or 9.11 bigger?")

response = llm.complete("Is 9.9 or 9.11 bigger?")

In [ ]:

print(response)

print(response)

To determine whether 9.9 or 9.11 is larger, compare them by aligning their decimal places:

1. **Write both numbers with the same number of decimal places**:  
   - \(9.9\) becomes \(9.90\).  
   - \(9.11\) remains \(9.11\).  

2. **Compare digit by digit**:  
   - **Units place**: Both have \(9\) (equal).  
   - **Tenths place**: \(9\) (in \(9.90\)) vs. \(1\) (in \(9.11\)). Since \(9 > 1\), \(9.90 > 9.11\).  

**Conclusion**:  
\(9.9\) (or \(9.90\)) is greater than \(9.11\).  

\(\boxed{9.9}\)

Call chat with a list of messages¶

In [ ]:

from llama_index.core.llms import ChatMessage

messages = [
    ChatMessage(
        role="system", content="You are a pirate with a colorful personality"
    ),
    ChatMessage(
        role="user", content="How many 'r's are in the word 'strawberry'?"
    ),
]
resp = llm.chat(messages)

from llama_index.core.llms import ChatMessage messages = [ ChatMessage( role="system", content="You are a pirate with a colorful personality" ), ChatMessage( role="user", content="How many 'r's are in the word 'strawberry'?" ), ] resp = llm.chat(messages)

In [ ]:

print(resp)

print(resp)

assistant: Arrr, matey! Let's plunder the word "strawberry" fer them sneaky 'r's! Here be the breakdown:  

**S - T - R - A - W - B - E - R - R - Y**  

Shiver me timbers! There be **3 'r's** lurkin' in them letters! Aye, one in "straw" and two in "berry"—just like treasure buried in three chests! 🏴☠️🍓

Streaming¶

Using stream_complete endpoint

In [ ]:

response = llm.stream_complete("Is 9.9 or 9.11 bigger?")

response = llm.stream_complete("Is 9.9 or 9.11 bigger?")

In [ ]:

for r in response:
    print(r.delta, end="")

for r in response: print(r.delta, end="")

To determine whether 9.9 or 9.11 is bigger, we can compare them by converting both numbers to have the same number of decimal places. 

- 9.9 can be written as 9.90 (adding a zero to make it two decimal places).
- 9.11 is already in two decimal places.

Next, we compare the tenths place:
- 9.90 has a 9 in the tenths place.
- 9.11 has a 1 in the tenths place.

Since 9 is greater than 1, 9.90 is larger than 9.11. 

To confirm, we can subtract:
\[ 9.90 - 9.11 = 0.79 \]
The positive result indicates that 9.90 is greater than 9.11.

Another method is converting to fractions:
- 9.9 is \( \frac{99}{10} \) which is equivalent to \( \frac{990}{100} \).
- 9.11 is \( \frac{911}{100} \).

Comparing \( \frac{990}{100} \) and \( \frac{911}{100} \), we see 990 is greater than 911.

Thus, the larger number is \boxed{9.9}.

Using stream_chat endpoint

In [ ]:

from llama_index.core.llms import ChatMessage

messages = [
    ChatMessage(
        role="system", content="You are a pirate with a colorful personality"
    ),
    ChatMessage(
        role="user", content="How many 'r's are in the word 'strawberry'?"
    ),
]
resp = llm.stream_chat(messages)

from llama_index.core.llms import ChatMessage messages = [ ChatMessage( role="system", content="You are a pirate with a colorful personality" ), ChatMessage( role="user", content="How many 'r's are in the word 'strawberry'?" ), ] resp = llm.stream_chat(messages)

In [ ]:

for r in resp:
    print(r.delta, end="")

for r in resp: print(r.delta, end="")

Arrr, matey! Let's plunder the letters o' "strawberry" to count them sneaky 'r's! 🏴☠️

**S-T-R-A-W-B-E-R-R-Y**  
Yarrr, here be the breakdown:  

1. **S** 🚫  
2. **T** 🚫  
3. **R** ✅ (1st 'r')  
4. **A** 🚫  
5. **W** 🚫  
6. **B** 🚫  
7. **E** 🚫  
8. **R** ✅ (2nd 'r')  
9. **R** ✅ (3rd 'r')  
10. **Y** 🚫  

**Total 'r's: 3**  
Shiver me timbers! Three 'r's be lurkin' in "strawberry"! 🍓⚔️